Rozprawy doktorskie: „Graphyne networks”

1

Xu, Zhen. "On-surface synthesis of two-dimensional graphene nanoribbon networks". Kyoto University, 2020. http://hdl.handle.net/2433/254529.

Pełny tekst źródła

Style APA, Harvard, Vancouver, ISO itp.

2

Garman, Paul Douglas. "Chemical Vapor Deposition of Silicon Oxycarbide Catalyzed Graphene Networks". The Ohio State University, 2018. http://rave.ohiolink.edu/etdc/view?acc_num=osu1523898208600691.

Pełny tekst źródła

Style APA, Harvard, Vancouver, ISO itp.

3

Ramli, Muhammad M. "Carbon nanotubes and graphene oxide networks for gas sensing". Thesis, University of Surrey, 2015. http://epubs.surrey.ac.uk/807845/.

Pełny tekst źródła

Streszczenie:

Nowadays, Carbon Nanotubes (CNTs) and Graphene Oxide (GO) have attracted enormous interest in various applications such as thin film transistors, chemical sensors, field emission devices and transparent conductive coatings. In this report, thin films of CNT and GO networks were fabricated as a two-terminal device for gas sensing applications. These devices were fabricated using vacuum filtration and drop casting methods at room temperature in order to get a thin and uniform film. The electrical measurement was conducted in order to investigate the film resistance as the solution concentration increased. The multi – walled carbon nanotubes (MWCNTs) solutions were acid treated by attaching carboxylic acid (-COOH) groups, in order to form a stable aqueous suspension with a neutral pH. The conductivity of the networks film was increased as the solution concentration increased where the sheet resistance at the highest concentration (0.125 mg/ml) was around 7.09 kΩ/sq. The single – walled carbon nanotubes (SWCNTs) suspension was produced using an organic dye in order to improve its solubility in water. Raman spectroscopy showed that no damage to the structure of SWCNTs was occurred. The GO suspension was produced by the chemical exfoliation of graphite through oxidation. The basal plane and edges of GO were decorated by oxygen functional groups, hence improving its solubility in water. To decrease the concentration of solutions, the solutions were diluted with methanol or water and various concentrations were achieved. The extreme sensitivity to changes in CNTs and GO local chemical environment makes them an ideal candidate for gas sensing application. The devices were tested by exposure to gasses such as NO2 and NH3. Results show a tremendous sensitivity towards NO2 and NH3 gasses. The sensitivities of the MWCNTs sensor device were ranging from 2 to 20 %. Whereas for the SWCNTs sensor device, the sensitivities were ranging from 20 to 50 %. For the GO sensor devices, the optimum sensitivity was achieved when the device was exposed at room temperature. The changes in resistance of the devices reflected the interaction mechanism that happened between exposure gasses (NO2 and NH3) and the materials.

Style APA, Harvard, Vancouver, ISO itp.

4

De, Marco Martina. "Hierarchical carbon nanotube and graphene oxide networks for multifunctional applications". Thesis, Imperial College London, 2016. http://hdl.handle.net/10044/1/47972.

Pełny tekst źródła

Streszczenie:

Assembling carbon nanomaterials (CNs) into networks and macrostructures is a potentially effective approach for the development of a wide array of technologies, including energy storage and production devices. CNs, such as carbon nanotubes (CNTs) and graphene (G), are characterised by impressive mechanical and electrical properties, however, these features are related to the high quality, individualised single carbon species.1, 2 Producing two/three-dimensional CN architectures presents several hurdles, mainly concerning the need to disassemble the pristine CN aggregates, the damages inflicted on the carbon framework during processing, and the consequent lack of mechanical strength and/or reduced electrical conductivity of the final material. Suitable methods for preparing CN (macro)structures retaining the extraordinary properties of the fundamental CN units, have yet to be fully developed. This Thesis addresses these issues by suggesting two different methodologies for the synthesis of CN networks, which are tailored to specific applications of the final structures. A novel cross-linking strategy of single-walled carbon nanotubes (SWCNTs) is developed, yielding highly connected, high surface area (> 750 m2 g-1) and electrically conductive (> 15 S m-1) cryogels. The cryogels are demonstrated to be effective electrodes within fully working electrochemical devices. In contrast to cross-linking strategies already explored in literature, the SWCNTs are individualised at high concentrations (up to 0.25 M), and cross-linked with p-diiodobenzene without shortening or damaging the carbon framework via a “reductive chemistry” route.3 Careful control of the absolute charge concentration in the system is found to be crucial for maximising the extent of debundling and grafting, with a suggested optimum at 15 mM. Optimised synthesis parameters in turn determine the accessible surface area and the conductive properties of the final freeze-dried cryogels. Multi-walled carbon nanotubes (MWCNTs) and graphene oxide (GO) hybrid networks are produced by a simple mixing approach, and used as supports for the CO2 adsorbents Layered Double Hydroxides (LDHs).4 Due to a strong synergistic interaction between the two CNs, hybrid GO/MWCNT systems significantly outperform the pure CNs as a support for LDHs, providing improved CO2 adsorption capacity and dramatically enhanced multicycle thermal stability (up to 96% of gas capacity retention after 20 cycles of adsorption-desorption). Detailed materials characterisation at several stages of the multicycle CO2 adsorption process, links the improved performance to the microstructure, showing that the hybrid GO/MWCNT substrate provides a superior surface area and LDH dispersion from the start, and resists sintering more effectively than either pure GO or pure MWCNTs. A systematic investigation of the relative proportions of the three-phase mixture (MWCNT/GO/LDH) consistently identifies a GO/MWCNT 1:1 ratio as the optimum for both surface area and sorbent performance, particularly when added in small amount (10 - 20%) to LDHs.

Style APA, Harvard, Vancouver, ISO itp.

5

Iqbal, Muhammad Zahir. "Structural and electrical characterization of doped graphene and carbon nanotube networks". Doctoral thesis, Universitat Politècnica de Catalunya, 2014. http://hdl.handle.net/10803/277570.

Pełny tekst źródła

Streszczenie:

Graphene, a two-dimensional sp2-hybridized network of carbon atoms has received a remarkable cornucopia of new physics and served as a unique model system, due particularly to its electronic properties, which could have interesting applications in electronic, spintronic or quantum devices. The first part of the thesis describes the modulation of graphene¿s structural and electrical properties with various kinds of doping; such as deep ultraviolet irradiation in ambient atmosphere, deep ultraviolet light irradiation in different gaseous environments, and electron beam irradiation. We have fabricated graphene (exfoliated and chemical vapor deposition grown graphene) field effect transistors using photolithography and electron beam lithography and characterized with AFM, Raman spectroscopy and transport measurement using low noise standard lock-in amplifier technique. We have explored how the ultraviolet light exposure tunes the electrical properties of graphene in an ambient atmosphere, confirmed by the shift of Dirac point position towards positive gate voltage, revealing p-type doping for graphene without degradation of mobility. We found that the doping is stable for a time scale of months. This method became more useful when half the graphene device was exposed by ultraviolet light, while the other half part was covered by a mask to make a sharp p-n junction. The doping effect became more prominent and controllable when it was made in an oxygen environment. The most interesting phenomena were observed when doped graphene was restored to a pristine state using ultraviolet light irradiation in a nitrogen environment. Furthermore, we have investigated the doping tunability with ultraviolet light irradiation on mechanically exfoliated single-, bi-, and trilayer graphenes without significantly degrading its charge carrier mobility. In a further study, the structural deformation of graphene was investigated by irradiation of an electron beam. The graphene structure changes its phase in various stages, where graphene transforms gradually from a crystalline to a nanocrystalline form and after a certain irradiation time into an amorphous form. This irradiation effect acts as an n-type dopant for graphene. In this case, mobility decreases with the gradual increase of irradiation dose, which implies the formation of localized states. The second part of the thesis describes carbon nanotube networks as flexible and transparent electrodes for electronic devices, particularly for high frequency applications. The observed results show that at low frequencies, the impedance increases as the density of nanotube networks decreases, as expected. Both the real and imaginary parts of impedance (measured up to 20 GHz) abruptly decrease as the frequency increases over the cut-off frequency. The cut-off frequency not only depends on the carbon nanotube density of the network, but also on the sample geometry. The Nyquist diagram suggests a simple equivalent circuit composed of a parallel combination of a resistor and a capacitor. The experimental results are in line with calculations made by electrochemical spectroscopy simulations. The results show that the electrical behavior is mostly determined by the contact resistance between the nanotubes, which are in a completely disordered distribution in the network. We show that carbon nanotube flexible conducting films, which may be transparent, could be competitive for some applications, such as displays, photovoltaic solar cells or selective sensors.
El grafè, considerat com una xarxa bidimensional d’àtoms de carboni units per enllaços híbrids sp2, és un tema de recerca molt prolífer en els últims anys, com a model de sòlid bidimensional, i molt particularment degut a les seves propietats electròniques, que poden tenir aplicacions interessants en dispositius electrònics, spintrònics o quàntics. La primera part de la Tesi descriu la modificació de les propietats estructurals i elèctriques del grafè utilitzant diferents mètodes per a dopar-lo: radiació ultraviolada d’alta energia (DUV) en atmosfera ambient, DUV en diferents gasos tals com oxigen o nitrogen, o irradiant amb un feix d’electrons (e-beam). Hem fabricat transistors d’efecte de camp (FET) amb grafè (exfoliat a partir del grafit, o bé obtingut per deposició química en fase vapor, CVD) utilitzant fotolitografia i e-beam litografia, i els hem caracteritzat mitjançant AFM, espectroscòpia Raman i mesures de transport elèctric, per a les que hem utilitzat la tècnica d’amplificació de baix soroll, el lock-in. Hem investigat com l’exposició a la llum ultraviolada en atmosfera ambient, modula les propietats elèctriques del grafè, de manera que la posició del punt de Dirac es desplaça cap a tensions de porta positives, cosa que implica dopatge de tipus-p, sense que hi hagi degradació de la mobilitat. El dopatge és estable al menys durant mesos. Amb el mateix mètode, quan només la meitat del dispositiu és exposat a la radiació ultraviolada mentre l’altre meitat és recobert per una màscara metàl·lica, hem obtingut una unió p-n. L’efecte de dopatge és més important i controlable, quan és fet en atmosfera d’oxigen. L’efecte més interessant que hem observat és la reversibilitat, quan el grafè dopat retorna al seu estat primitiu, en ser irradiat amb llum ultraviolada en atmosfera de nitrogen. També hem investigat el dopatge amb llum ultraviolada del grafè exfoliat mecànicament, de una, dues o tres capes, observant que es produeix sense una degradació significativa de la mobilitat dels portadors de càrrega. Posteriorment hem estudiat la deformació estructural del grafè quan és irradiat amb un feix d’electrons. Hem observat canvis estructurals en diferents etapes: el grafè evoluciona gradualment, a partir de la forma cristal·lina, cap a una fase d’estructura nanocristal·lina i finalment, després d’una certa dosi de irradiació, presenta una estructura amorfa. L’efecte d’ irradiar el grafè amb electrons actua com a dopant tipus-n, però en aquest cas la mobilitat decreix en incrementar la dosi, això implica que hi ha formació d’estats localitzats. La segona part de la Tesi tracta de capes primes de nanotubs de carboni, com a elèctrodes flexibles i transparents per a dispositius electrònics, en particular per aplicacions d’alta freqüència. Els resultats obtinguts mostren que, a baixes freqüències, la impedància augmenta en disminuir la densitat de nanotubs, tal com cal esperar. Tan la part real com la part imaginària de la impedància (mesurada fins a 20 GHz) decreixen abruptament en augmentar la freqüència més enllà de la freqüència de tall. La freqüència de tall no depèn únicament de la densitat de nanotubs en la capa, sinó també de la geometria de la mostra. El diagrama de Nyquist es pot interpretar amb un circuit equivalent consistent simplement en una resistència i un condensador en paral·lel. Els resultats experimentals s’ajusten bé a les simulacions fetes per espectroscòpia d’impedàncies (EIS). Els resultats posen en evidència que el comportament elèctric queda majoritàriament determinat per la resistència de contacte entre els nanotubs, que formen la xarxa amb una distribució totalment desordenada. Hem vist que capes primes de nanotubs de carboni conductores i flexibles, que poden ser també transparents, poden ser competitives en diferents aplicacions, com ara pantalles, cel·les solars fotovoltaiques o sensors selectius

Style APA, Harvard, Vancouver, ISO itp.

6

MARTIN, JIMENEZ CRISTINA. "Nuovi materiali compositi basati su grafene". Doctoral thesis, Università degli Studi di Trieste, 2016. http://hdl.handle.net/11368/2908013.

Pełny tekst źródła

Streszczenie:

The number of possibilities and, hence, examples of composite materials, is huge and varied and depend on the nature of the matrix. However, the evolution of polymers since the beginning of the 19th Century has led to a boost in research on that field, in which the latest advances have provided numerous advantages in nanomedicine, for instance. As a matter of fact, in recent years composites have been made from materials in which the size of at least one of the phases is in the nanometer scale, also called nanomaterials. Among all of them, graphene has unique properties owing to its sp2 hybridised carbon atoms arranged in a 2D honeycomb lattice. Therefore, the incorporation of graphene into polymeric materials offers new options regarding the use of these eventual composites in a wider range of fields (from sensors to biological applications). In this context, the aim of this thesis is the design of new graphene-based composites to be used in several applications regarding the design and the specific features of the final prepared materials. Chapter 1 provides a revision of composite materials, primarily describing important concepts for the understanding of this thesis such as hydrogels, interpenetrating polymer networks (IPNs), and, undoubtedly, graphene; and digging down to more specific literature examples mainly based on this fascinating nanomaterial. In Chapter 2, new composite hydrogels with an autonomous self-healing capacity are described. Semi-IPNs of poly (methacrylic acid) (pMAAc) and poly (vinyl alcohol) are primarily studied. The materials are firstly synthesised and characterised in the presence or absence of graphene, and their healing abilities are subsequently analysed. They show not only an electromechanical behaviour, but also an almost complete auto-reparation after being damaged. Interestingly, graphene does not limit any of the studied properties. The second type of materials were prepared from the MAAc in combination with [2-(acryloyloxy)ethyl]trimethylammonium chloride. The repairing capacity is studied depending on the way in which the material is prepared, and also depending on the presence or not of graphene. In this case, the self-healing is based on ionic interactions. All the materials have demonstrated good repairing ability, even with graphene. Chapter 3 details the synthesis, characterization and different applications of graphene-based composite hydrogels using acrylamide as the main monomer. The composites show excellent mechanical properties, and they are even proved successfully as 3D scaffolds for cell culture, being possible nucleus pulposus replacements for intervertebral disc diseases. De de-swelling behaviour under an external stimulus is also demonstrated, making these composites applicable as probable on-demand drug delivery systems. Finally, a piezoresistive effect is detailed in this chapter due to the presence of graphene into the hydrogel network, obtaining excellent gauge factor values because of the change in the resistivity of the material depending on whether it is being stretched or not. That fact pares the way for the use of our systems as strain sensors. Finally, in Chapter 4, the creation of novel nanostructured systems for neural network regeneration are presented. Firstly, the graphene addition on 2D substrates previously coated with single-wall carbon nanotubes does not show a notable enhancement in the neuron activity, obtaining, for instance, similar values for membrane capacitance with respect to the substrate only coated with the carbon nanotubes. In a second approach, neuron activity is studied in a graphene-based 3D scaffold. The most important fact observed in this study is that neurons are only visualised in the graphene-based hydrogel, but not in the one without nanomaterial, confirming that graphene is taking an important role in the neuronal network.

Style APA, Harvard, Vancouver, ISO itp.

7

Yarmolenko, O. V., S. A. Baskakov, Y. M. Shulga, P. I. Vengrus i O. N. Efimov. "Supercapacitors Based on Composite Polyaniline / Reduced Graphene Oxide with Network Nanocomposite Polymer Electrolyte". Thesis, Sumy State University, 2013. http://essuir.sumdu.edu.ua/handle/123456789/35510.

Pełny tekst źródła

Streszczenie:

The paper describes investigation on new types of supercapacitors based on composite polyani-line/reduced graphene oxide with network nanocomposite polymer electrolyte. Its prototypes are all solid state. The new network polymer electrolytes based on polyethylene glycol diacrylate and nanoparticle SiO2 was synthesized by reaction of radical polymerization in the environment of liquid organic electrolyte. The work is aimed to obtain a polymer electrolyte that is compatible with the electrode materials of superca-pacitors. For these purposes the method of FTIR spectroscopy, a.c. electrochemical impedance and gal-vanostatic cycling were used. When you are citing the document, use the following link http://essuir.sumdu.edu.ua/handle/123456789/35510

Style APA, Harvard, Vancouver, ISO itp.

8

Bloess, Mark. "Harnessing Social Networks for Social Awareness via Mobile Face Recognition". Thèse, Université d'Ottawa / University of Ottawa, 2013. http://hdl.handle.net/10393/23792.

Pełny tekst źródła

Streszczenie:

With more and more images being uploaded to social networks each day, the resources for identifying a large portion of the world are available. However the tools to harness and utilize this information are not sufficient. This thesis presents a system, called PhacePhinder, which can build a face database from a social network and have it accessible from mobile devices. Through combining existing technologies, this is made possible. It also makes use of a fusion probabilistic latent semantic analysis to determine strong connections between users and content. Using this information we can determine the most meaningful social connection to a recognized person, allowing us to inform the user of how they know the person being recognized. We conduct a series of offline and user tests to verify our results and compare them to existing algorithms. We show, that through combining a user’s friendship information as well as picture occurrence information, we can make stronger recommendations than based on friendship alone. We demonstrate a working prototype that can identify a face from a picture taken from a mobile phone, using a database derived from images gathered directly from a social network, and return a meaningful social connection to the recognized face.

Style APA, Harvard, Vancouver, ISO itp.

9

EL, MERHIE AMIRA. "Single Layer Graphene Biointerface: Studying Neuronal Network Development and Monitoring Cell Behavior over Time". Doctoral thesis, Università degli studi di Genova, 2019. http://hdl.handle.net/11567/939896.

Pełny tekst źródła

Streszczenie:

The objective of my Ph.D. thesis is the investigation of the role of Single Layer Graphene (SLG) as a biointerface for its possible future exploitation in various biomedical applications; in particular for the development of biosensors, substrates for regenerative medicine, interfacing platforms for better recording of electrophysiological activity of neuronal networks, among others. This Ph.D. project is multidisciplinary involving both the material transfer and characterization part from one side and the biological part from another side. The material part offers an in-depth explanation of SLG synthesis, transfer, characterization and functionalization while the biological section sheds light on the studies performed for investigation of the behavior of different types of cell lines on SLG substrates. For better understanding of the sequence of the performed work, I have divided this thesis into separate chapters. In the beginning and end of every chapter, I added an introduction and conclusions related to it. Chapter 1 acts as a general introduction to graphene and graphene-related materials where a detailed explanation on the evolution of those materials as a cell interface is provided leading to the introduction of SLG in the end of this chapter along with its production process. Chapter 2 is oriented on the surface characterization of SLG substrates; in this chapter, I described the SLG transfer method, creation of the micrometric ablated geometric patterns on the transferred substrates using excimer laser micromachining, a technique developed in our lab, then further functionalization of the substrates and finally all the techniques employed for their physicochemical characterization. Chapter 3 is dedicated to the biological part of the project; i.e. studying the behavior of different cell lines on the SLG substrates. In this chapter, I have described and explained the interest of using the selected cell lines and the experiments that were performed on them. Chapter 4 has been devoted to a complete and separate project that I performed in collaboration with the Neuroscience and Brain Technologies department. The main focus of the project was the functionalization of the commercial multi-electrode arrays (MEAs) with SLG and studying the neuronal network activity on them throughout the complete network development. Although the main focus of my Ph.D. project was studying SLG biointerface, I have also been involved in side projects, among which, studying the neuronal-like response of mouse neuroblastoma (N2a) living cells to nanoporous patterns of thin supported anodic alumina which I have described in Appendix A, and studying the surface potential of graphene by polyelectrolyte coating which I have presented in Appendix B. To summarize, this thesis reports an original investigation, since, to the best of our knowledge, there is no report yet about the study of the effect of SLG functionalized MEA on the neuronal network activity throughout the complete network maturation. Furthermore, proliferation curves of different cell lines on SLG versus control substrates have been presented; in addition to physicochemical characterization of ablated and functionalized SLG substrates as means of possible explanation of a certain cellular behavior on graphene.

Style APA, Harvard, Vancouver, ISO itp.

10

Ren, Haolin. "Visualizing media with interactive multiplex networks". Thesis, Bordeaux, 2019. http://www.theses.fr/2019BORD0036/document.

Pełny tekst źródła

Streszczenie:

Les flux d’information suivent aujourd’hui des chemins complexes: la propagation des informations, impliquant éditeurs on-line, chaînes d’information en continu et réseaux sociaux, emprunte alors des chemins croisés, susceptibles d’agir sur le contenu et sa perception. Ce projet de thèse étudie l’adaptation des mesures de graphes classiques aux graphes multiplexes en relation avec le domaine étudié, propose de construire des visualisations à partir de plusieurs représentations graphiques des réseaux, et de les combiner (visualisations multi-vues synchronisées, représentations hybrides, etc.). L’accent est mis sur les modes d’interaction permettant de prendre en compte l’aspect multiplexe (multicouche) des réseaux. Ces représentations et manipulations interactives s’appuient aussi sur le calcul d’indicateurs propres aux réseaux multiplexes. Ce travail est basé sur deux jeux de données principaux: l’un est une archive de 12 ans de l’émission japonaise publique quotidienne NHK News 7, de 2001 à 2013. L’autre recense les participants aux émissions de télévision/radio françaises entre 2010 et 2015. Deux systèmes de visualisation s’appuyant sur une interface Web ont été développés pour analyser des réseaux multiplexes, que nous appelons «Visual Cloud» et «Laputa». Dans le Visual Cloud, nous définissons formellement une notion de similitude entre les concepts et les groupes de concepts que nous nommons possibilité de co-occurrence (CP). Conformément à cette définition, nous proposons un algorithme de classification hiérarchique. Nous regroupons les couches dans le réseau multiplexe de documents, et intégrons cette hiérarchie dans un nuage de mots interactif. Nous améliorons les algorithmes traditionnels de disposition de mise en forme de nuages de mots de sorte à préserver les contraintes sur la hiérarchie de concepts. Le système Laputa est destiné à l’analyse complexe de réseaux temporels denses et multidimensionnels. Pour ce faire, il associe un graphe à une segmentation. La segmentation par communauté, par attribut, ou encore par tranche temporelle, forme des vues de ce graphe. Afin d’associer ces vues avec le tout global, nous utilisons des diagrammes de Sankey pour révéler l’évolution des communautés (diagrammes que nous avons augmentés avec un zoom sémantique). Cette thèse nous permet ainsi de parcourir trois aspects (3V) des plus intéressants de la donnée et du BigData appliqués aux archives multimédia: Le Volume de nos données dans l’immensité des archives, nous atteignons des ordres de grandeurs qui ne sont pas praticables pour la visualisation et l’exploitation des liens. La Vélocité à cause de la nature temporelle de nos données (par définition). La Variété qui est un corollaire de la richesse des données multimédia et de tout ce que l’on peut souhaiter vouloir y investiguer. Ce que l’on peut retenir de cette thèse c’est que la traduction de ces trois défis a pris dans tous les cas une réponse sous la forme d’une analyse de réseaux multiplexes. Nous retrouvons toujours ces structures au coeur de notre travail, que ce soit de manière plus discrète dans les critères pour filtrer les arêtes par l’algorithme Simmelian backbone, que ce soit par la superposition de tranches temporelles, ou bien que ce soit beaucoup plus directement dans la combinaison d’indices sémantiques visuels et textuels pour laquelle nous extrayons les hiérarchies permettant notre visualisation
Nowadays, information follows complex paths: information propagation involving on-line editors, 24-hour news providers and social medias following entangled paths acting on information content and perception. This thesis studies the adaptation of classical graph measurements to multiplex graphs, to build visualizations from several graphical representations of the networks, and to combine them (synchronized multi-view visualizations, hybrid representations, etc.). Emphasis is placed on the modes of interaction allowing to take in hand the multiplex nature (multilayer) of the networks. These representations and interactive manipulations are also based on the calculation of indicators specific to multiplex networks. The work is based on two main datasets: one is a 12-year archive of the Japanese public daily broadcast NHK News 7, from 2001 to 2013. Another lists the participants in the French TV/radio shows between 2010 and 2015. Two visualization systems based on a Web interface have been developed for multiplex network analysis, which we call "Visual Cloud" and "Laputa". In the Visual Cloud, we formally define a notion of similarity between concepts and groups of concepts that we call co-occurrence possibility (CP). According to this definition, we propose a hierarchical classification algorithm. We aggregate the layers in a multiplex network of documents, and integrate that hierarchy into an interactive word cloud. Here we improve the traditional word cloud layout algorithms so as to preserve the constraints on the concept hierarchy. The Laputa system is intended for the complex analysis of dense and multidimensional temporal networks. To do this, it associates a graph with a segmentation. The segmentation by communities, by attributes, or by time slices, forms views of this graph. In order to associate these views with the global whole, we use Sankey diagrams to reveal the evolution of the communities (diagrams that we have increased with a semantic zoom). This thesis allows us to browse three aspects of the most interesting aspects of the data miming and BigData applied to multimedia archives: The Volume since our archives are immense and reach orders of magnitude that are usually not practicable for the visualization; Velocity, because of the temporal nature of our data (by definition). The Variety that is a corollary of the richness of multimedia data and of all that one may wish to want to investigate. What we can retain from this thesis is that we met each of these three challenges by taking an answer in the form of a multiplex network analysis. These structures are always at the heart of our work, whether in the criteria for filtering edges using the Simmelian backbone algorithm, or in the superposition of time slices in the complex networks, or much more directly in the combinations of visual and textual semantic indices for which we extract hierarchies allowing our visualization

Style APA, Harvard, Vancouver, ISO itp.

11

Zhou, Hang. "Graph algorithms : network inference and planar graph optimization". Thesis, Paris, Ecole normale supérieure, 2015. http://www.theses.fr/2015ENSU0016/document.

Pełny tekst źródła

Streszczenie:

Cette thèse porte sur deux sujets d’algorithmique des graphes. Le premier sujet est l’inférence de réseaux. Quelle est la complexité pour déterminer un graphe inconnu à partir de requêtes de plus court chemin entre ses sommets ? Nous supposons que le graphe est de degré borné. Dans le problème de reconstruction, le but est de reconstruire le graphe ; tandis que dans le problème de vérification, le but est de vérifier qu’un graphe donné est correct. Nous développons des algorithmes probabilistes utilisant une décomposition en cellules de Voronoi. Ensuite, nous analysons des algorithmes de type glouton, et montrons qu’ils sont quasi-optimaux. Nous étudions aussi ces problèmes sur des familles particulières de graphes, démontrons des bornes inférieures, et étudions la reconstruction approximative. Le deuxième sujet est l’étude de deux problèmes d’optimisation sur les graphes planaires. Dans le problème de classification par corrélations, l’entrée est un graphe pondéré, où chaque arête a une étiquette h+i ou h-i, indiquant si ses extrémités sont ou non dans la même catégorie. Le but est de trouver une partition des sommets en catégories qui respecte au mieux les étiquettes. Dans le problème d’augmentation 2-arête-connexe, l’entrée est un graphe pondéré et un sous-ensemble R des arêtes. Le but est de trouver un sous-ensemble S des arêtes de poids minimum, tel que pour chaque arête de R, ses extrémités sont dans une composante 2-arête-connexe de l’union de R et S. Pour les graphes planaires, nous réduisons le premier problème au deuxième et montrons que les deux problèmes, bien que NP-durs, ont un schéma d’approximation en temps polynomial. Nous utilisons la technique récente de décomposition en briques
This thesis focuses on two topics of graph algorithms. The first topic is network inference. How efficiently can we find an unknown graph using shortest path queries between its vertices? We assume that the graph has bounded degree. In the reconstruction problem, the goal is to find the graph; and in the verification problem, the goal is to check whether a given graph is correct. We provide randomized algorithms based on a Voronoi cell decomposition. Next, we analyze greedy algorithms, and show that they are near-optimal. We also study the problems on special graph classes, prove lower bounds, and study the approximate reconstruction. The second topic is optimization in planar graphs. We study two problems. In the correlation clustering problem, the input is a weighted graph, where every edge has a label of h+i or h−i, indicating whether its endpoints are in the same category or in different categories. The goal is to find a partition of the vertices into categories that tries to respect the labels. In the two-edge-connected augmentation problem, the input is a weighted graph and a subset R of edges. The goal is to produce a minimum-weight subset S of edges, such that for every edge in R, its endpoints are two-edge-connected in the union of R and S. For planar graphs, we reduce correlation clustering to two-edge-connected augmentation, and show that both problems, although they are NP-hard, have a polynomial-time approximation scheme. We build on the brick decomposition technique developed recently

Style APA, Harvard, Vancouver, ISO itp.

12

Yeliyur, Siddegowda Darshan. "Gray-cast iron classification based on graphite flakes using image morphology and neural networks". Thesis, California State University, Long Beach, 2016. http://pqdtopen.proquest.com/#viewpdf?dispub=10017846.

Pełny tekst źródła

Streszczenie:

Gray-cast iron is an iron carbon alloy which is regularly used in manufacturing processes. Carbon is distributed in the iron material in the form of graphite. The distribution of the graphite flakes in the alloy contributes greatly towards the chemical and physical properties of the metal alloy. Thus it is important to identify and classify the Gray-cast iron based on the morphological parameters of the graphite flakes. Gray-Cast iron is classified into five types in ISO-945 represented with the letters A through E. These five classes possess different structures or distributions of the graphite flakes. The current project presents an automated classification method using image processing and machine learning algorithms. The method presented here obtains the required parameters from the microstructure through image morphological operations. The image information is subsequently fed through a supervised machine learning algorithm which is trained using parameters such as area of the flakes, perimeter, minimum inter-particle distance and chord length from over twenty samples. The algorithm calculates the percentage of the type of the flakes present in the given image. The simulation is done in MATLAB and was tested for six images in each class. Class C and D were classified with 100 percent accuracy, Class A and B were classified with accuracy of 82 percent and Class E was identified with accuracy of 68 percent.

Style APA, Harvard, Vancouver, ISO itp.

13

Celik, Numan. "Wireless graphene-based electrocardiogram (ECG) sensor including multiple physiological measurement system". Thesis, Brunel University, 2017. http://bura.brunel.ac.uk/handle/2438/15698.

Pełny tekst źródła

Streszczenie:

In this thesis, a novel graphene (GN) based electrocardiogram (ECG) sensor is designed, constructed and tested to validate the concept of coating GN, which is a highly electrically conductive material, on Ag substrates of conventional electrodes. The background theory, design, experiments and results for the proposed GN-based ECG sensor are also presented. Due to the attractive electrical and physical characteristics of graphene, a new ECG sensor was investigated by coating GN onto itself. The main focus of this project was to examine the effect of GN on ECG monitoring and to compare its performance with conventional methods. A thorough investigation into GN synthesis on Ag substrate was conducted, which was accompanied by extensive simulation and experimentation. A GN-enabled ECG electrode was characterised by Raman spectroscopy, scanning electron microscopy along with electrical resistivity and conductivity measurements. The results obtained from the GN characteristic experimentation on Raman spectroscopy, detected a 2D peak in the GN-coated electrode, which was not observed with the conventional Ag/AgCl electrode. SEM characterisation also revealed that a GN coating smooths the surface of the electrode and hence, improves the skin-to-electrode contact. Furthermore, a comparison regarding the electrical conductivity calculation was made between the proposed GN-coated electrodes and conventional Ag/AgCl ones. The resistance values obtained were 212.4 Ω and 28.3 Ω for bare and GN-coated electrodes, respectively. That indicates that the electrical conductivity of GN-based electrodes is superior and hence, it is concluded that skin-electrode contact impedance can be lowered by their usage. Additional COMSOL simulation was carried out to observe the effect of an electrical field and surface charge density using GN-coated and conventional Ag/AgCl electrodes on a simplified human skin model. The results demonstrated the effectiveness of the addition of electrical field and surface charge capabilities and hence, coating GN on Ag substrates was validated through this simulation. This novel ECG electrode was tested with various types of electrodes on ten different subjects in order to analyse the obtained ECG signals. The experimental results clearly showed that the proposed GN-based electrode exhibits the best performance in terms of ECG signal quality, detection of critical waves of ECG morphology (P-wave, QRS complex and T-wave), signal-to-noise ratio (SNR) with 27.0 dB and skin-electrode contact impedance (65.82 kΩ at 20 Hz) when compared to those obtained by conventional a Ag/AgCl electrode. Moreover, this proposed GN-based ECG sensor was integrated with core body temperature (CBT) sensor in an ear-based device, which was designed and printed using 3D technology. Subsequently, a finger clipped photoplethysmography (PPG) sensor was integrated with the two-sensors in an Arduino based data acquisition system, which was placed on the subject's arm to enable a wearable multiple physiological measurement system. The physiological information of ECG and CBT was obtained from the ear of the subject, whilst the PPG signal was acquired from the finger. Furthermore, this multiple physiological signal was wirelessly transmitted to the smartphone to achieve continuous and real-time monitoring of physiological signals (ECG, CBT and PPG) on a dedicated app developed using the Java programming language. The proposed system has plenty of room for performance improvement and future development will make it adaptabadaptable, hence being more convenient for the users to implement other applications than at present.

Style APA, Harvard, Vancouver, ISO itp.

14

Wan, William Bradley. "The synthesis of fused and derivatized dehydrobenzoannulenes and hexakis(phenylbutadiynyl)benzenes for elucidating the aromatic and NLO properties of the all-carbon network, graphdiyne /". view abstract or download file of text, 2001. http://wwwlib.umi.com/cr/uoregon/fullcit?p3018400.

Pełny tekst źródła

Streszczenie:

Thesis (Ph. D.)--University of Oregon, 2001.
Typescript. Includes vita and abstract. Includes bibliographical references (leaves 237-258). Also available for download via the World Wide Web; free to University of Oregon users.

Style APA, Harvard, Vancouver, ISO itp.

15

Friggeri, Adrien. "A Quantitative Theory of Social Cohesion". Phd thesis, Ecole normale supérieure de lyon - ENS LYON, 2012. http://tel.archives-ouvertes.fr/tel-00737199.

Pełny tekst źródła

Streszczenie:

Community, a notion transversal to all areas of Social Network Analysis, has drawn tremendous amount of attention across the sciences in the past decades. Numerous attempts to characterize both the sociological embodiment of the concept as well as its observable structural manifestation in the social network have to this date only converged in spirit. No formal consensus has been reached on the quantifiable aspects of community, despite it being deeply linked to topological and dynamic aspects of the underlying social network. Presenting a fresh approach to the evaluation of communities, this thesis introduces and builds upon the cohesion, a novel metric which captures the intrinsic quality, as a community, of a set of nodes in a network. The cohesion, defined in terms of social triads, was found to be highly correlated to the subjective perception of communitiness through the use of a large-scale online experiment in which users were able to compute and rate the quality of their social groups on Facebook. Adequately reflecting the complexity of social interactions, the problem of finding a maximally cohesive group inside a given social network is shown to be NP-hard. Using a heuristic approximation algorithm, applications of the cohesion to broadly different use cases are highlighted, ranging from its application to network visualization, to the study of the evolution of agreement groups in the United States Senate, to the understanding of the intertwinement between subjects' psychological traits and the cohesive structures in their social neighborhood. The use of the cohesion proves invaluable in that it offers non-trivial insights on the network structure and its relation to the associated semantic.

Style APA, Harvard, Vancouver, ISO itp.

16

Khames, Imene. "Nonlinear network wave equations : periodic solutions and graph characterizations". Thesis, Normandie, 2018. http://www.theses.fr/2018NORMIR04/document.

Pełny tekst źródła

Streszczenie:

Dans cette thèse, nous étudions les équations d’ondes non-linéaires discrètes dans des réseaux finis arbitraires. C’est un modèle général, où le Laplacien continu est remplacé par le Laplacien de graphe. Nous considérons une telle équation d’onde avec une non-linéarité cubique sur les nœuds du graphe, qui est le modèle φ4 discret, décrivant un réseau mécanique d’oscillateurs non-linéaires couplés ou un réseau électrique où les composantes sont des diodes ou des jonctions Josephson. L’équation d’onde linéaire est bien comprise en termes de modes normaux, ce sont des solutions périodiques associées aux vecteurs propres du Laplacien de graphe. Notre premier objectif est d’étudier la continuation des modes normaux dans le régime non-linéaire et le couplage des modes en présence de la non-linéarité. En inspectant les modes normaux du Laplacien de graphe, nous identifions ceux qui peuvent être étendus à des orbites périodiques non-linéaires. Il s’agit des modes normaux dont les vecteurs propres du Laplacien sont composés uniquement de {1}, {-1,+1} ou {-1,0,+1}. Nous effectuons systématiquement une analyse de stabilité linéaire (Floquet) de ces orbites et montrons le couplage des modes lorsque l’orbite est instable. Ensuite, nous caractérisons tous les graphes pour lesquels il existe des vecteurs propres du Laplacien ayant tous leurs composantes dans {-1,+1} ou {-1,0,+1}, en utilisant la théorie spectrale des graphes. Dans la deuxième partie, nous étudions des solutions périodiques localisées spatialement. En supposant une condition initiale de grande amplitude localisée sur un nœud du graphe, nous approchons l’évolution du système par l’équation de Duffing pour le nœud excité et un système linéaire forcé pour le reste du réseau. Cette approximation est validée en réduisant l’équation φ4 discrète à l’équation de Schrödinger non-linéaire de graphes et par l’analyse de Fourier de la solution numérique. Les résultats de cette thèse relient la dynamique non-linéaire à la théorie spectrale des graphes
In this thesis, we study the discrete nonlinear wave equations in arbitrary finite networks. This is a general model, where the usual continuum Laplacian is replaced by the graph Laplacian. We consider such a wave equation with a cubic on-site nonlinearity which is the discrete φ4 model, describing a mechanical network of coupled nonlinear oscillators or an electrical network where the components are diodes or Josephson junctions. The linear graph wave equation is well understood in terms of normal modes, these are periodic solutions associated to the eigenvectors of the graph Laplacian. Our first goal is to investigate the continuation of normal modes in the nonlinear regime and the modes coupling in the presence of nonlinearity. By inspecting the normal modes of the graph Laplacian, we identify which ones can be extended into nonlinear periodic orbits. They are normal modes whose Laplacian eigenvectors are composed uniquely of {1}, {-1,+1} or {-1,0,+1}. We perform a systematic linear stability (Floquet) analysis of these orbits and show the modes coupling when the orbit is unstable. Then, we characterize all graphs for which there are eigenvectors of the graph Laplacian having all their components in {-1,+1} or {-1,0,+1}, using graph spectral theory. In the second part, we investigate periodic solutions that are spatially localized. Assuming a large amplitude localized initial condition on one node of the graph, we approximate its evolution by the Duffing equation. The rest of the network satisfies a linear system forced by the excited node. This approximation is validated by reducing the discrete φ4 equation to the graph nonlinear Schrödinger equation and by Fourier analysis. The results of this thesis relate nonlinear dynamics to graph spectral theory

Style APA, Harvard, Vancouver, ISO itp.

17

Labarre, Anthony. "Combinatorial aspects of genome rearrangements and haplotype networks". Phd thesis, Universite Libre de Bruxelles, 2008. http://tel.archives-ouvertes.fr/tel-00482196.

Pełny tekst źródła

Streszczenie:

La thèse couvre deux problèmes motivés par la biologie: l'étude des réarrangements génomiques, et celle des réseaux d'haplotypes. Les problèmes de réarrangements génomiques sont un cas particulier des problèmes de distances d'édition, où l'on cherche à transformer un objet en un autre en utilisant le plus petit nombre possible d'opérations, les opérations autorisées étant fixées au préalable; on s'intéresse également à la distance entre les deux objets, c'est-à-dire au calcul du nombre d'opérations dans une séquence optimale plutôt qu'à la recherche d'une telle séquence. Les problèmes de réarrangements génomiques peuvent souvent s'exprimer comme des problèmes de tri de permutations (vues comme des arrangements linéaires de {1,2,...,n}) en utilisant le plus petit nombre d'opérations (autorisées) possible. Nous examinons en particulier les ``transpositions', qui déplacent un intervalle de la permutation. Beaucoup de problèmes liés au tri par transpositions sont ouverts, en particulier sa complexité algorithmique. Nous nous écartons des ``outils standards' utilisés dans le domaine des réarrangements génomiques, et utilisons la décomposition en cycles disjoints des permutations pour prouver de nouvelles majorations sur la distance des transpositions ainsi que des formules permettant de calculer cette distance en temps polynomial dans de nombreux cas. Cette décomposition nous sert également à résoudre un problème d'énumération concernant le ``graphe des cycles' de Bafna et Pevzner, et à construire une technique générale permettant d'obtenir de nouvelles minorations en reformulant tous les problèmes de distances d'édition sur les permutations en termes de factorisations de permutations paires associées. Les réseaux d'haplotypes sont des graphes dont une partie des sommets porte des étiquettes, utilisés en génomique comparative quand les arbres sont trop restrictifs, ou quand l'on ne peut choisir une ``meilleure' topologie parmi un ensemble donné d'arbres. Nous formalisons une nouvelle méthode due à Cassens, Mardulyn et Milinkovitch, qui consiste à construire un graphe contenant tous les arbres partiellement étiquetés donnés et possédant le moins d'arêtes possible, et donnons des algorithmes résolvant le problème de manière optimale sur deux graphes, dont le temps d'exécution est exponentiel en général mais polynomial dans quelques cas que nous caractérisons.

Style APA, Harvard, Vancouver, ISO itp.

18

Ruiz, Mathieu. "Codage cortical de la synesthésie graphème-couleur". Thesis, Grenoble, 2014. http://www.theses.fr/2014GRENS020/document.

Pełny tekst źródła

Streszczenie:

La synesthésie est un phénomène fascinant qui offre une opportunité privilégiée d'étudier chez les sujets sains les bases neurales de l'expérience subjective. Les synesthètes graphème-couleur (1 à 5 % de la population – qui ignorent le plus souvent l'être) associent arbitrairement et systématiquement une couleur spécifique à chaque lettre ou chiffre. Ce travail de thèse se focalise sur ces synesthésies et explore l'implication, dans l'expérience subjective de couleurs synesthésiques, des régions corticales qui sont actives lors de la perception des couleurs ‘réelles'. Dans une étude préalable réalisé par l'équipe d'accueil utilisant l'Imagerie par Résonance Magnétique (IRM) fonctionnelle, une telle implication des ‘aires de la couleur' n'a pas été observée (Hupé et al. 2012). Une analyse statistique ‘classique' (univariée) du traitement des données mesurées était utilisée.Cette thèse vise à déterminer si les couleurs synesthésiques reposent sur les réseaux neuronaux des couleurs réelles en utilisant une technique d'analyse statistique multivariée (MultiVoxel Pattern Analysis – MVPA). A la différence de l'analyse univariée elle se base sur des ensembles de voxels (les pixels en 3D qui composent les images acquises) et prend en compte leurs motifs d'activation correspondant spécifiquement à l'encodage d'une information donnée par le cerveau. Cet encodage est réalisé au niveau neuronal et l'IRMf en donne une quantification non-invasive et indirecte au travers des répercutions hémodynamiques induites par l'activation neuronale. Cette modélisation multivariée des données fait des MVPA une approche particulièrement adaptée à la mesure d'informations encodées finement de manière distribuée. Le but de ce travail est d'explorer son efficacité pour l'étude des synesthésies graphème-couleur où l'analyse classique n'a pas fourni de résultats robustes. En pratique, elle nécessite cependant l'utilisation de protocoles spécifiques, la maîtrise des nombreux paramètres qui influent radicalement sur son fonctionnement et l'utilisation conjointe de l'analyse univariée. Dans une première étape de cette thèse, nous avons évalué différents aspects méthodologiques qu'il est important de maîtriser afin d'obtenir des résultats robustes et une analyse fiable.Ensuite, nous avons comparé le traitement des couleurs réelles et synesthésiques chez deux groupes de 20 synesthètes et non synesthètes. Nous avons trouvé que le traitement des couleurs synesthésiques ne repose pas sur les réseaux neuronaux de traitement de la couleur réelle. Cela peut signifier que les bases neuronales des couleurs synesthésiques ne se situent ni dans les aires visuelles rétinotopiques ni dans les aires de l'expertise visuelle (les aires de la ‘voie ventrale'). Cela peut également signifier que, bien que ces aires soient impliquées, ce ne sont pas les réseaux neuronaux de traitement des couleurs réelles qui sont activés lors de la perception de couleurs synesthésique. D'un point de vue méthodologique, il est possible que les signaux mesurés par l'IRMf ne permettent pas d'observer ce codage partagé. Ces résultats posent donc la question des limites de l'interprétation des signaux mesurés en IRMf très indirectement liés à l'activité neuronale. L'identification des réseaux neuronaux impliqués dans l'expérience subjective des couleurs synesthésiques reste donc une problématique ouverte
Synesthesia is a fascinating phenomenon that offers the opportunity to study the neural bases of subjective experiences in healthy subjects. Grapheme-color synesthetes (1 to 5 % of the population – who do not know it most of the time) arbitrarily and systematically associate a specific color to letters or digits. This PhD thesis work focuses on this type of synesthesia and explores whether common neural networks are involved both in ‘real' color perception and synesthetic color experience. In a previous study from the host team using functional Magnetic Resonance Imaging (MRI), no implication of ‘color areas' where found (Hupé et al., 2012). A standard (univariate) statistical analysis of the data processing was used.This PhD thesis aims at determining if synesthetic colors involve real color neural networks with the use of a multivariate statistical technique (Multivoxel Pattern Analysis – MVPA). Unlike univariate analysis it uses sets of voxels (the pixels in 3D forming the images) and take into account their patterns of activation linked to the encoding of specific information in the brain. This encoding is performed at the neuronal level and fMRI indirectly and non-invasively quantifies it through hemodynamic variations induced by the neuronal activity. MVPA is a particularly adapted approach to measure fine grained and distributed information encoding. The goal of the thesis is to explore its efficiency for the study of grapheme-color synesthesia for which standard analyses failed. In practice, it requires the use of specific protocols, mastering numerous parameters influencing the results and the joint use of univariate analysis. In the first step of this thesis, we evaluated different methodological aspects to optimize the processing chain in order to obtain robust and reliable results.Then, we compared the neural processing of real colors and synesthetic colors in 2 groups of synesthetes (n=20) and non synesthetes (n=20). We found that synesthetic colors processing does not share common neural networks with real color processing. This suggests that the neural bases of synesthetic colors are not localized in the retinotopic visual areas or in the visual expertise areas (the ‘ventral pathway' areas). This may also suggest that, although those areas are involved, different neural networks are implicated in real color and synesthetic color perception. These results raise the question of the limits of the interpretation of the signal measured by fMRI, indirectly linked to the neuronal activity. The identification of the neural networks involved in the subjective experience of synesthetic colors remains an open issue

Style APA, Harvard, Vancouver, ISO itp.

19

Samavat, Reza. "Mean Eigenvalue Counting Function Bound for Laplacians on Random Networks". Doctoral thesis, Universitätsbibliothek Chemnitz, 2015. http://nbn-resolving.de/urn:nbn:de:bsz:ch1-qucosa-159578.

Pełny tekst źródła

Streszczenie:

Spectral graph theory widely increases the interests in not only discovering new properties of well known graphs but also proving the well known properties for the new type of graphs. In fact all spectral properties of proverbial graphs are not acknowledged to us and in other hand due to the structure of nature, new classes of graphs are required to explain the phenomena around us and the spectral properties of these graphs can tell us more about the structure of them. These both themes are the body of our work here. We introduce here three models of random graphs and show that the eigenvalue counting function of Laplacians on these graphs has exponential decay bound. Since our methods heavily depend on the first nonzero eigenvalue of Laplacian, we study also this eigenvalue for the graph in both random and nonrandom cases.

Style APA, Harvard, Vancouver, ISO itp.

20

Pasdeloup, Bastien. "Extending convolutional neural networks to irregular domains through graph inference". Thesis, Ecole nationale supérieure Mines-Télécom Atlantique Bretagne Pays de la Loire, 2017. http://www.theses.fr/2017IMTA0048/document.

Pełny tekst źródła

Streszczenie:

Tout d'abord, nous présentons des méthodes permettant d'inférer un graphe à partir de signaux, afin de modéliser le support des données à classifier. Ensuite, des translations préservant les voisinages des sommets sont identifiées sur le graphe inféré. Enfin, ces translations sont utilisées pour déplacer un noyau convolutif sur le graphe, afin dedéfinir un réseau de neurones convolutif adapté aux données d'entrée.Nous avons illustré notre méthodologie sur une base de données d'images. Sans utiliser de connaissances sur les signaux, nous avons pu inférer un graphe proche d'une grille. Les translations sur ce graphe sont proches des translations Euclidiennes, ce qui nous a permis de définir un réseau de neurones convolutif très similaire à ce que l'on aurait pu obtenir en utilisant l'information que les signaux sont des images. Ce réseau, entraîné sur les données initiales, a dépassé lesperformances des méthodes de l'état de l'art de plus de 13 points, tout en étant simple et facilement améliorable.La méthode que nous avons introduite est une généralisation des réseaux de neurones convolutifs, car ceux-ci sont des cas particuliers de notre approche quand le graphe est une grille. Nos travaux ouvrent donc de nombreuses perspectives, car ils fournissent une méthode efficace pour construire des réseaux adaptés aux données
This manuscript sums up our work on extending convolutional neuralnetworks to irregular domains through graph inference. It consists of three main chapters, each giving the details of a part of a methodology allowing the definition of such networks to process signals evolving on graphs with unknown structures.First, graph inference from data is explored, in order to provide a graph modeling the support of the signals to classify. Second, translation operators that preserve neighborhood properties of the vertices are identified on the inferred graph. Third, these translations are used to shift a convolutional kernel on the graph in order to define a convolutional neural network that is adapted to the input data.We have illustrated our methodology on a dataset of images. While not using any particular knowledge on the signals, we have been able to infer a graph that is close to a grid. Translations on this graph resemble Euclidean translations. Therefore, this has allowed us to define an adapted convolutional neural network that is very close what one would obtain when using the information that signals are images. This network, trained on the initial data, has out performed state of the art methods by more than 13 points, while using a very simple and easily improvable architecture.The method we have introduced is a generalization of convolutional neural networks. As a matter of fact, they can be seen as aparticularization of our approach in the case where the graph is a grid. Our work thus opens the way to numerous perspectives, as it provides an efficient way to build networks that are adapted to the data

Style APA, Harvard, Vancouver, ISO itp.

21

Leger, Jean-Benoist. "Modelling the topology of ecological bipartite networks with statistical models for heterogeneous random graphs". Paris 7, 2014. http://www.theses.fr/2014PA077185.

Pełny tekst źródła

Streszczenie:

Un réseau écologique constitue une représentation de l'ensemble des interactions entre espèces dans un contexte donné. L'analyse de la structure topologique de ces réseaux permet aux écologues d'identifier et de comprendre les processus sous-jacents. La détection de sous-groupes d'espèces interagissant fortement ensemble, souvent nommés communautés o compartiments, est un des principaux moyens pour interpréter la structure sous-jacente des réseaux. Il existe de nombreuses méthodes de classification non supervisée, qui peuvent être utilisées pour analyser des réseaux écologiques. L'analyse des réseaux est actuellement un sujet de recherche en pleine expansion avec des applications dans des domaines de recherches variés. À notre connaissance, il n'existe pas de comparaison des méthodes de classification non supervisée dans le cas des réseaux écologique. Nous avons effectué une revue des méthodes disponibles de classification non supervisées des noeuds, et nous les avons comparé dans un contexte écologique. Afin d'évaluer la contribution des différents processus expliquant la structure d'un réseau, nous avons introduit de l'information extérieure au réseau. Nous avons analysé deux réseaux d'interaction arbre-champignon et arbre-insecte. Ces résultats sont préliminaires, mais la méthode semble ouvrir des perspectives intéressantes en écologie. Nous avons également étudié un réseau écologique de nature différente, un réseau de reproduction entre arbres. Nous avons utilisé ces résultats pour discuter d'un concept central en écologie, le concept d'espèce
An ecological network is a representation of the whole set of interactions between species in a given context. Ecological scientists analyse the topological structure of such networks, in order to understand the underlying processes. The identification of sub-groups of highly-interacting species (usually called communities, or compartments) is an important stream of research. The most popular method for the search of communities in ecological networks is the modularity optimization method. However this popularity is more due to the first paper published on this topic than to a rational choice based on solid grounds. There are many other clustering methods that could be used to delimit communities in ecological networks. The analysis of complex networks is indeed a rapidly growing topic with many applications in several scientific fields. To our knowledge, no comparison of different clustering methods is available in the case of ecological networks. Here we reviewed the whole set of methods available for clustering networks and we compared them using an ecological benchmark. In order to assess the relative contribution of several processes to the network structure, we integrated exogenous information in the clustering model. We analysed two bipartite antagonistic networks with this method, a tree-fungus and tree-insect network. The results are still preliminary but the method seems to us very promising for future ecological studies. Finally we searched communities in a different kind of network, a mating network between individuals belonging to two hybridizing tree species. We used our results to discuss a concept which is central in ecology, the species concept

Style APA, Harvard, Vancouver, ISO itp.

22

Peng, Han. "Spatial resolved electronic structure of low dimensional materials and data analysis". Thesis, University of Oxford, 2018. http://ora.ox.ac.uk/objects/uuid:2f3503eb-93bf-48d6-b6fb-13409b925748.

Pełny tekst źródła

Streszczenie:

Two dimensional (2D) materials with interesting fundamental physics and potential applications attract tremendous efforts to study. The versatile properties of 2D materials can be further tailored by tuning the electronic structure with the layer-stacking arrangement, of which the main adjustable parameters include the thickness and the in-plane twist angle between layers. The Angle-Resolved Photoemission Spectroscopy (ARPES) has become a canonical tool to study the electronic structure of crystalline materials. The recent development of ARPES with sub-micrometre spatial resolution (micro-ARPES) has made it possible to study the electronic structure of materials with mesoscopic domains. In this thesis, we use micro-ARPES to investigate the spatially-resolved electronic structure of a series of few-layer materials: 1. We explore the electronic structure of the domains with different number of layers in few-layer graphene on copper substrate. We observe a layer- dependent substrate doping effect in which the Fermi surface of graphene shifts with the increase of number of layers, which is then explained by a multilayer effective capacitor model. 2. We systematically study the twist angle evolution of the energy band of twisted few-layer graphene over a wide range of twist angles (from 5° to 31°). We directly observe van Hove Singularities (vHSs) in twisted bilayer graphene with wide tunable energy range over 2 eV. In addition, the formation of multiple vHSs (at different binding energies) is observed in trilayer graphene. The large tuning range of vHS binding energy in twisted few-layer graphene provides a promising material base for optoelectrical applications with broad-band wavelength selectivity. 3. To better extract the energy band features from ARPES data, we propose a new method with a convolutional neural network (CNN) that achieves comparable or better results than traditional derivative based methods. Besides ARPES study, this thesis also includes the study of surface reconstruction for the layered material Bi2O2Se with the analysis of Scanning Tunnelling Microscopy (STM) images. To explain the origin of the pattern, we propose a tile model that produces the identical statistics with the experiment.

Style APA, Harvard, Vancouver, ISO itp.

23

Riediger, Steffen. "Implementierung eines Algorithmus zur Partitionierung von Graphen". Thesis, Universitätsbibliothek Chemnitz, 2007. http://nbn-resolving.de/urn:nbn:de:swb:ch1-200701098.

Pełny tekst źródła

Streszczenie:

Partitionierung von Graphen ist im Allgemeinen sehr schwierig. Es stehen derzeit keine Algorithmen zur Verfügung, die ein allgemeines Partitionierungsproblem effizient lösen. Aus diesem Grund werden heuristische Ansätze verfolgt. Zur Analyse dieser Heuristiken ist man derzeit gezwungen zufällige Graphen zu Verwenden. Daten realer Graphen sind derzeit entweder nur sehr schwer zu erheben (z.B. Internetgraph), oder aus rechtlichen bzw. wirtschaftlichen Gründen nicht zugänglich (z.B. soziale Netzwerke). Die untersuchten Heuristiken liefern teilweise nur unter bestimmten Voraussetzungen Ergebnisse. Einige arbeiten lediglich auf einer eingeschränkten Menge von Graphen, andere benötigen zum Erkennen einer Partition einen mit der Knotenzahl steigenden Durchschnittsgrad der Knoten, z.B. [DHM04]. Der im Zuge dieser Arbeit erstmals implementierte Algorithmus aus [CGL07a] benötigt lediglich einen konstanten Durchschnittsgrad der Knoten um eine Partition des Graphen, wenn diese existiert, zu erkennen. Insbesondere muss dieser Durchschnittsgrad nicht mit der Knotenzahl steigen. Nach der Implementierung erfolgten Tests des Algorithmus an zufälligen Graphen. Diese Graphen entsprachen dem Gnp-Modell mit eingepflanzter Partition. Die untersuchten Clusterprobleme waren dabei große Schnitte, kleine Schnitte und unabhängige Mengen. Der von der Art des Clusterproblems abhängige Durchschnittsgrad wurde während der Tests bestimmt.

Style APA, Harvard, Vancouver, ISO itp.

24

Ezzeddine, Diala. "A contribution to topological learning and its application in Social Networks". Thesis, Lyon 2, 2014. http://www.theses.fr/2014LYO22011/document.

Pełny tekst źródła

Streszczenie:

L'Apprentissage Supervisé est un domaine populaire de l'Apprentissage Automatique en progrès constant depuis plusieurs années. De nombreuses techniques ont été développées pour résoudre le problème de classification, mais, dans la plupart des cas, ces méthodes se basent sur la présence et le nombre de points d'une classe donnée dans des zones de l'espace que doit définir le classifieur. Á cause de cela la construction de ce classifieur est dépendante de la densité du nuage de points des données de départ. Dans cette thèse, nous montrons qu'utiliser la topologie des données peut être une bonne alternative lors de la construction des classifieurs. Pour cela, nous proposons d'utiliser les graphes topologiques comme le Graphe de Gabriel (GG) ou le Graphes des Voisins Relatifs (RNG). Ces dernier représentent la topologie de données car ils sont basées sur la notion de voisinages et ne sont pas dépendant de la densité. Pour appliquer ce concept, nous créons une nouvelle méthode appelée Classification aléatoire par Voisinages (Random Neighborhood Classification (RNC)). Cette méthode utilise des graphes topologiques pour construire des classifieurs. De plus, comme une Méthodes Ensemble (EM), elle utilise plusieurs classifieurs pour extraire toutes les informations pertinentes des données. Les EM sont bien connues dans l'Apprentissage Automatique. Elles génèrent de nombreux classifieurs à partir des données, puis agrègent ces classifieurs en un seul. Le classifieur global obtenu est reconnu pour être très eficace, ce qui a été montré dans de nombreuses études. Cela est possible car il s'appuie sur des informations obtenues auprès de chaque classifieur qui le compose. Nous avons comparé RNC à d'autres méthodes de classification supervisées connues sur des données issues du référentiel UCI Irvine. Nous constatons que RNC fonctionne bien par rapport aux meilleurs d'entre elles, telles que les Forêts Aléatoires (RF) et Support Vector Machines (SVM). La plupart du temps, RNC se classe parmi les trois premières méthodes en terme d'eficacité. Ce résultat nous a encouragé à étudier RNC sur des données réelles comme les tweets. Twitter est un réseau social de micro-blogging. Il est particulièrement utile pour étudier l'opinion à propos de l'actualité et sur tout sujet, en particulier la politique. Cependant, l'extraction de l'opinion politique depuis Twitter pose des défis particuliers. En effet, la taille des messages, le niveau de langage utilisé et ambiguïté des messages rend très diffcile d'utiliser les outils classiques d'analyse de texte basés sur des calculs de fréquence de mots ou des analyses en profondeur de phrases. C'est cela qui a motivé cette étude. Nous proposons d'étudier les couples auteur/sujet pour classer le tweet en fonction de l'opinion de son auteur à propos d'un politicien (un sujet du tweet). Nous proposons une procédure qui porte sur l'identification de ces opinions. Nous pensons que les tweets expriment rarement une opinion objective sur telle ou telle action d'un homme politique mais plus souvent une conviction profonde de son auteur à propos d'un mouvement politique. Détecter l'opinion de quelques auteurs nous permet ensuite d'utiliser la similitude dans les termes employés par les autres pour retrouver ces convictions à plus grande échelle. Cette procédure à 2 étapes, tout d'abord identifier l'opinion de quelques couples de manière semi-automatique afin de constituer un référentiel, puis ensuite d'utiliser l'ensemble des tweets d'un couple (tous les tweets d'un auteur mentionnant un politicien) pour les comparer avec ceux du référentiel. L'Apprentissage Topologique semble être un domaine très intéressant à étudier, en particulier pour résoudre les problèmes de classification
Supervised Learning is a popular field of Machine Learning that has made recent progress. In particular, many methods and procedures have been developed to solve the classification problem. Most classical methods in Supervised Learning use the density estimation of data to construct their classifiers.In this dissertation, we show that the topology of data can be a good alternative in constructing classifiers. We propose using topological graphs like Gabriel graphs (GG) and Relative Neighborhood Graphs (RNG) that can build the topology of data based on its neighborhood structure. To apply this concept, we create a new method called Random Neighborhood Classification (RNC).In this method, we use topological graphs to construct classifiers and then apply Ensemble Methods (EM) to get all relevant information from the data. EM is well known in Machine Learning, generates many classifiers from data and then aggregates these classifiers into one. Aggregate classifiers have been shown to be very efficient in many studies, because it leverages relevant and effective information from each generated classifier. We first compare RNC to other known classification methods using data from the UCI Irvine repository. We find that RNC works very well compared to very efficient methods such as Random Forests and Support Vector Machines. Most of the time, it ranks in the top three methods in efficiency. This result has encouraged us to study the efficiency of RNC on real data like tweets. Twitter, a microblogging Social Network, is especially useful to mine opinion on current affairs and topics that span the range of human interest, including politics. Mining political opinion from Twitter poses peculiar challenges such as the versatility of the authors when they express their political view, that motivate this study. We define a new attribute, called couple, that will be very helpful in the process to study the tweets opinion. A couple is an author that talk about a politician. We propose a new procedure that focuses on identifying the opinion on tweet using couples. We think that focusing on the couples's opinion expressed by several tweets can overcome the problems of analysing each single tweet. This approach can be useful to avoid the versatility, language ambiguity and many other artifacts that are easy to understand for a human being but not automatically for a machine.We use classical Machine Learning techniques like KNN, Random Forests (RF) and also our method RNC. We proceed in two steps : First, we build a reference set of classified couples using Naive Bayes. We also apply a second alternative method to Naive method, sampling plan procedure, to compare and evaluate the results of Naive method. Second, we evaluate the performance of this approach using proximity measures in order to use RNC, RF and KNN. The expirements used are based on real data of tweets from the French presidential election in 2012. The results show that this approach works well and that RNC performs very good in order to classify opinion in tweets.Topological Learning seems to be very intersting field to study, in particular to address the classification problem. Many concepts to get informations from topological graphs need to analyse like the ones described by Aupetit, M. in his work (2005). Our work show that Topological Learning can be an effective way to perform classification problem

Style APA, Harvard, Vancouver, ISO itp.

25

MacGill, Ross. "Fast moving neutrons, graphite moderators and radioactive clouds : an ANT account of the Chernobyl accident's risky network". Thesis, University of Glasgow, 2017. http://theses.gla.ac.uk/8260/.

Pełny tekst źródła

Streszczenie:

This research seeks to understand the Chernobyl accident at the material level and this is achieved through a sustained engagement and use of Latour’s Actor-Network Theory. The primary research question asks how has the accident travelled through various worlds: impacting upon them, reorienting their structure, and at times, creating new nuclear worlds in its wake. The ambiguous nature of the question is designed to allow the research to approach the expansive topic from several angles, thus accounting for the very real material spread radioactive fallout and, crucially, the material action engendered by the agency of RBMK no4, the central actant in the network in question. The nature of this work is quasi-scientific; that is, the majority of the empirical data is the discourses of the ‘hard sciences’, the disciplines tasked with understanding and mitigating one of the worst accidents the world has ever witnessed, and safeguarding against another similar occurrence. The reassembling of RBMK no4’s network, and its travels, hence ‘renders visible’ the lessons learnt by the nuclear industry in the wake of the event of April 1986. The spectre of Chernobyl has loomed over the industry for 3 decades. This research attempts to create a pragmatic context. It is not interested in the epistemological ‘sensationalist’ representations of the accident. Instead it embraces the complex intricacies of the physics, engineering and radiological sciences — the very essence of the materiality of RBMK no4, and its risky network.

Style APA, Harvard, Vancouver, ISO itp.

26

Wychowaniec, Jacek. "Designing nanostructured peptide hydrogels containing graphene oxide and its derivatives for tissue engineering and biomedical applications". Thesis, University of Manchester, 2018. https://www.research.manchester.ac.uk/portal/en/theses/designing-nanostructured-peptide-hydrogels-containing-graphene-oxide-and-its-derivatives-for-tissue-engineering-and-biomedical-applications(409e60a2-ed17-45bf-ab6c-b76ede937a67).html.

Pełny tekst źródła

Streszczenie:

Progress in biomedicine requires the design of functional biomaterials, in particular, 3-dimensional (3D) scaffolds. Shear thinning, β-sheet based peptide hydrogels have attracted wide interest due to their potential use in tissue engineering and biomedical applications as 3D functional scaffolds. The emergence of carbon nanomaterials has also opened the door for the construction of increasingly functional hybrid hydrogels built from nanofibres and graphene-based materials using non-covalent physical interactions. The relationship between peptide molecular structure and the formed hydrogel is important for understanding the material response to shear. In particular, the physicochemical properties of peptide based biomaterials will affect the feasibility of injecting them during medical procedures. In the first part of this work, four peptides: FEFKFEFK (F8), FKFEFKFK (FK), KFEFKFEFK (KF8) and KFEFKFEFKK (KF8K) (F - phenylalanine, E - glutamic acid, K - lysine) were designed and used at identical charge to explore the effect of lysine rich β-sheet self-assembling sequences on the shear thinning behaviour and final properties of bulk hydrogels. By varying the peptide sequence design and concentration of the peptide, the tendency of the nanofibres formed to aggregate and the balance of nanofibre junction strength versus fibre cohesive strength could be explored. This allowed the existing theory of the shear thinning behaviour of this class of materials to be extended. The relationship between molecular structures of nanofibres forming the 3D network and the nano-filler is critical to understand in order to design tuneable and functional materials. In the next part of the work, three rationally designed β-sheet peptides, which form hydrogels: VEVKVEVK (V8), FEFKFEFK (F8) and FEFEFKFE (FE) (V - valine) and five graphene-based materials: graphene oxide (GO), reduced graphene oxide (rGO), three graphene-polymer hybrid flakes: GO with polydiallyldimethylammonium chloride (GO/PDADMAC), rGO with PDADMAC (rGO/PDADMAC) and rGO with polyvinylpyrrolidone (rGO/PVP) were used to form a selection of hybrid hydrogels. Graphene derivatives of the lateral flake sizes of 16.8 ± 10.1 µm were used. Various interactions between the graphene flakes and the peptides were observed that affected the overall mechanical properties of the hydrogels. Electrostatic interactions and pie-pie stacking, when phenylalanine residues are present, were shown to play a key role in determining the dispersion of graphene materials in the peptide hydrogels and stiffness of the hybrid materials. In particular, FE with reduced graphene oxide (rGO) and FE with rGO covered with polydiallyldimethylammonium chloride (PDADMAC) thin film formed double network-like hybrid hydrogels due to strong formation of peptide nanofibrillar bridges between adjacent rGO flakes. This corresponded to the 3- and 4-fold increase in the storage modulus (Gꞌ) of these hydrogels in comparison to controls. FE hydrogels with homogeneus dispersions of graphene oxide (GO) and reduced graphene oxide (rGO) are further shown to be suitable for 3D culture of human mesenchymal stem cells (hMSCs) with no cytotoxicity. These results focus attention on the importance of understanding interactions between the nano-filler and the nanofibrillar network in forming hybrid hydrogels with tuneable mechanical and biological properties, and demonstrates the possibility of using these materials as 3D cell culture scaffolds for biomedical purposes. Furthermore, graphene oxide (GO) itself is currently used in a number of processes of technological relevance such as wet spinning, injection moulding or inkjet printing to form graphene fibres, composites and printed conductors. Typically, such processes utilise well-aligned layered GO liquid crystal (LC) structures in aqueous dispersions. Flow and confinement encountered during processing affects the alignment and stability of this phase. In the final part of this work, the alignment of GOLCs of two lateral flake sizes (42.1 ± 29.4 µm and 15.5 ± 7.5 µm) were probed under a wide range of rotational shear flow conditions that overlap with the manufacturing processes defined by angular speeds from 0.08 to 8 rad.s-1 (and corresponding maximum shear rates from 0.1 s-1 to 100 s-1), in real-time, using shear induced polarized light imaging and small angle X-ray scattering, both coupled with an in-situ rheometer (Rheo-SIPLI and Rheo-SAXS, respectively). Under certain conditions, a unique pattern in Rheo-SIPLI: a Maltese cross combined with shear banding was observed. This phenomenon is unique to GO flakes of sufficiently large lateral size. The structure formed is attributed to a helical flow arising from a combination of shear flow and Taylor-vortex type flow, which is reinforced by a mathematical model. The orientations prescribed by this model are consistent with anomalous rheopecty oberved in Rheo-SIPLI and an anomolous scattering pattern in Rheo-SAXS. With the current trend towards producing ultra-large GO flakes, evidence that the flow behaviour changes from a Couette flow to a Taylor vortex flow was provided, which would lead to undesired, or alternatively, controllable alignment of GO flakes for a variety of applications, including aligned structures for biomedical purposes.

Style APA, Harvard, Vancouver, ISO itp.

27

Kormoš, Lukáš. "2D molekulární systémy na površích". Doctoral thesis, Vysoké učení technické v Brně. CEITEC VUT, 2021. http://www.nusl.cz/ntk/nusl-438774.

Pełny tekst źródła

Streszczenie:

Molekulárne systémy predstavujú jeden zo smerov súčasného výskumu nových nanoelektronických zariadení. Organické molekuly nachádzajú uplatnenie v rôznych aplikáciách, ako sú napríklad solárne články, displeje alebo kvantové počítače. Rast vysokokvalitných molekulárnych vrstiev s požadovanými vlastnosťami často vyžaduje využitie samousporiadavaných štruktúr, hlboké pochopenie rozhrania kovu a organických molekúl a tiež dynamiky rastu molekulárnych vrstiev. Predkladaná práca sa zaoberá predovšetkým samousporadanými štruktúrami bifenyl-dikarboxylovej kyseliny (BDA) na Cu (1 0 0) a Ag (1 0 0), ktoré boli skúmané v UHV s využitím STM, XPS a LEEM. V prípade BDA-Ag je podrobne opísaných niekoľko chemicky a štrukturálne odlišných molekulárnych fáz. Ďalej boli BDA a TCNQ molekuly skúmané na grafene pripravenom na Ir (1 1 1). Okrem toho sa organo-kovové systémy syntetizovali depozíciou atómov Ni a Fe s molekulami TCNQ a BDA. Záverečná časť tejto práce popisuje povrchovú syntézu grafénových nanoribonov (7-AGNR) na špecicky štrukturovanom substráte Au (16 14 15) z prekurzorových molekúl DBBA. Rekonštrukcia povrchu po raste bola analyzovaná pomocou STM a elektronické vlastnosti 7-AGNRs pomocou ARPES.

Style APA, Harvard, Vancouver, ISO itp.

28

DEL, PUPPO SIMONE. "MATERIALI 2D SU SUPERFICI METALLICHE: UN APPROCCIO NUMERICO". Doctoral thesis, Università degli Studi di Trieste, 2023. https://hdl.handle.net/11368/3041020.

Pełny tekst źródła

Streszczenie:

Nonostante l'ampio lavoro già fatto sui materiali 2D,non si è ancora arrivati a una completa comprensione dell'interazione strato-substrato e di come ciò influisce sulla struttura e sulle proprietà elettroniche. Con l'obiettivo di far luce su questa questione ancora aperta, in questo lavoro sono stati investigati attraverso simulazioni numeriche alcuni sistemi basati su due diversi materiali 2D, Grafene (G) su substrati di Nichel (Ni)(la maggior parte del lavoro) e Blue-Phosphorus (BP) su substrati di Oro (Au), utilizzando due diversi approcci numerici. La maggior parte del lavoro consiste in simulazioni ab-initio basate sulla Teoria del funzionale densità(DFT), prestando attenzione ad alcuni dettagli tecnici specifici per garantire l'accuratezza e attendibilità dei risultati. Una parte del lavoro in questione invece riguarda la costruzione e validazione tramite tecniche di reti neurali di un nuovo potenziale interatomico, con l’obiettivo di estendere lo studio dei sistemi G/Ni a configurazioni più realistiche o a processi non affrontabili tramite calcoli ab-initio. Durante tutto il lavoro, un confronto diretto con risultati sperimentali,pubblicati o nuovi,sarà discusso. Il grafene può essere facilmente cresciuto mediante CVD su substrati di nichel, ma le sue proprietà elettroniche e strutturali dipendono dal matching/mismatching e dall'allineamento/disallineamento tra il suo reticolo esagonale e il reticolo superficiale sottostante. Nella prima parte di questo lavoro di tesi viene discusso il G epitassiale cresciuto su Ni(111), un sistema già noto in letteratura. Partendo da una serie di calcoli DFT, abbiamo utilizzato una rete neurale per generare un potenziale interatomico in grado di prevedere con precisione l'energia e le forze in questo sistema. Il nuovo potenziale consente di eseguire simulazioni di dinamica molecolare con migliaia di atomi con una precisione vicina a quella del DFT, aprendo la strada a simulazioni su larga scala per questi sistemi. Successivamente, è stata studiata la ricostruzione strutturale subita dalla superficie di Ni(111) ad alte temperature durante il processo CVD. In questo lavoro si mostra come la presenza di domini ruotati di G rispetto al reticolo Ni(111) influisce sulla formazione di una fase di carburo di nichel,Ni2C, sottostante. Inoltre, abbiamo studiato l'intercalazione del monossido di carbonio all’interfaccia G epitassiale e Ni(111), indagando in modo sistematico il pattern formato dal CO intercalato, evidenziando le modificazioni indotte sulla struttura elettronica del grafene. La traccia più importante dell'intercalazione del CO è uno spostamento dei coni di Dirac, linearmente dipendente dalla quantità di CO intercalata. In questo lavoro è stato studiato anche il G su Ni(100). Tale interfaccia, a causa del mismatch reticolare, presenta un pattern di moirè a strisce (stripè moirè) in cui regioni di G fortemente e debolmente interagenti con il substrato di Ni sottostante si alternano, inducendo proprietà elettroniche anisotrope e modulate e proprietà di reattività. In questo caso sono stati studiati diversi tipi di difetti dello strato di G e come questi possano aumentare la reattività del grafene per l’assorbimento di adatomi o molecole. L'ultima parte della tesi è stata dedicata al Blu-Fosforo, un nuovo materiale 2D composto da soli atomi di P disposti in modo simile al grafene ma con un parametro reticolare più grande e con gli atomi non giacenti sullo stesso piano. Il miglior modello individuato per descrivere il BP su Au(111) è formato da P9 domini di forma piramidale collegati tra loro da adatomi di Au in una supercella 5x5. La corrispondenza con le immagini sperimentali STM e gli spettri ARPES, consente di discriminare tra diversi modelli possibili, indicando ancora una volta la necessità di uno sforzo sinergico tra simulazioni ed esperimenti per far luce sulla struttura e sulle proprietà dei sistemi reali.
Although the extensive work on 2D materials, a comprehensive understanding of the layer-substrate interaction and how this affects the structure and the electronic properties is still lacking.With the aim of shedding light on this still open issue, in this work we extensively investigated through numerical simulations some selected systems based on two different 2D materials, Graphene (G) on Nickel (Ni) substrates (most of the work) and Blue-Phosphorus (BP) on Gold (Au) substrate, using different and, to some extent, complementary numerical approaches. Most of the work consisted in quantum mechanical ab-initio simulations based on Density Functional Theory (DFT), paying attention to some specific technical details to ensure the accuracy and the reliability of the results. Part of the work concerned the construction via neural network techniques and the validation of new interatomic potentials to extend the investigation of G/Ni systems to more realistic configurations or to dynamical processes not directly affordable by ab-initio calculations. Throughout the work, a direct comparison with published or new experimental results is discussed. Graphene can be easily grown by CVD on nickel substrates but its electronic and structural properties depend on the matching/mismatching and on the alignment/misalignment between its hexagonal lattice and the underlying surface lattice.The thesis starts with the investigation of epitaxial G on Ni(111), which is already very well known.Starting from a set of DFT calculations that we also used as a benchmark to refine many technical details of our simulations on other new configurations, we used a Neural Network to generate an interatomic potential able to accurately predict energy and forces in this system. The new potential allows to perform molecular dynamics simulations with thousands of atoms with accuracy close to that of DFT, paving the way for large-scale simulations of such system. We report a successful application on large G domains showing cohexistence of different registries with the substrate.After that, structural reconstruction that Ni(111) surface undergoes at high temperatures during CVD process has been investigated. We showed how the presence of rotated domains of graphene with respect to Ni(111) lattice affects the formation of a nickel carbide phase, Ni2C,underneath. Furthermore, we studied the intercalation of Carbon Monoxide under epitaxial G grown on Ni(111) providing a systematic investigation of the intercalated CO pattern, highlighting the modifications induced on the graphene electronic structure.The most important signature of CO intercalation is a shift of Dirac cones linearly dependent on the CO coverage, opening the way to application as gas sensor to easily detect and quantify its presence. In this work, G on Ni(100)has also been studied. Such an interface, due to lattice mismatch, presents a stripe moiré pattern in which strongly (chemisorbed) and weakly (physisorbed) interacting G regions with Ni surface alternate, inducing anisotropic modulated electronic structure and reactivity properties.Here we provided a full investigation of different kind of defects of G layer and we investigated how they can increase the reactivity of graphene for metal adatoms or molecules adsorption. The last part of the thesis has been devoted to Blue-Phosphorus,a new 2D material made by only P atoms arranged similarly to graphene but with a larger lattice parameter and a small buckling of the two constituent sublattices.To describe BP grown on Au(111),we identified as the best candidate a structural model formed by P9 pyramidal shaped domains connected by Au adatoms in a 5x5 supercell.The nice correspondence with experimental STM images and ARPES spectra, allows to discriminate among different possible models,indicating once again the necessity of a synergetic effort between simulations and experiments to shed light on the structure and properties of real systems.

Style APA, Harvard, Vancouver, ISO itp.

29

Moscu, Mircea. "Inférence distribuée de topologie de graphe à partir de flots de données". Thesis, Université Côte d'Azur, 2020. http://www.theses.fr/2020COAZ4081.

Pełny tekst źródła

Streszczenie:

La deuxième décennie du millénaire actuel peut être résumée en une courte phrase : l'essor des données. Le nombre de sources de données s'est multiplié : du streaming audio-vidéo aux réseaux sociaux et à l'Internet des Objets, en passant par les montres intelligentes, les équipements industriels et les véhicules personnels, pour n'en citer que quelques-unes. Le plus souvent, ces sources forment des réseaux afin d'échanger des informations. En conséquence directe, le domaine du Traitement de Signal sur Graphe a prospéré et a évolué. Son but : traiter et donner un sens à tout le déluge de données environnant. Dans ce contexte, le but principal de cette thèse est de développer des méthodes et des algorithmes capables d'utiliser des flots de données, de manière distribuée, afin d'inférer les réseaux sous-jacents qui relient ces flots. Ensuite, ces topologies de réseau estimées peuvent être utilisées avec des outils développés pour le Traitement de Signal sur Graphe afin de traiter et d'analyser les données supportées par des graphes. Après une brève introduction suivie d'exemples motivants, nous développons et proposons d'abord un algorithme en ligne, distribué et adaptatif pour l'inférence de topologies de graphes pour les flots de données qui sont linéairement dépendants. Une analyse de la méthode s'ensuit, afin d'établir des relations entre les performances et les paramètres nécessaires à l'algorithme. Nous menons ensuite une série d'expériences afin de valider l'analyse et de comparer ses performances avec celles d'une autre méthode proposée dans la littérature. La contribution suivante est un algorithme doté des mêmes capacités en ligne, distribuées et adaptatives, mais adapté à l'inférence de liens entre des données qui interagissent de manière non-linéaire. À ce titre, nous proposons un modèle additif simple mais efficace qui utilise l'usine du noyau reproduisant afin de modéliser lesdites non-linéarités. Les résultats de son analyse sont convaincants, tandis que les expériences menées sur des données biomédicales donnent des réseaux estimés qui présentent un comportement prédit par la littérature médicale. Enfin, une troisième proposition d'algorithme est faite, qui vise à améliorer le modèle non-linéaire en lui permettant d'échapper aux contraintes induites par l'additivité. Ainsi, le nouveau modèle proposé est aussi général que possible, et utilise une manière naturelle et intuitive d'imposer la parcimonie des liens, basée sur le concept de dérivés partiels. Nous analysons également l'algorithme proposé, afin d'établir les conditions de stabilité et les relations entre ses paramètres et ses performances. Une série d'expériences est menée, montrant comment le modèle général est capable de mieux saisir les liens non-linéaires entre les données, tandis que les réseaux estimés se comportent de manière cohérente avec les estimations précédentes
The second decade of the current millennium can be summarized in one short phrase: the advent of data. There has been a surge in the number of data sources: from audio-video streaming, social networks and the Internet of Things, to smartwatches, industrial equipment and personal vehicles, just to name a few. More often than not, these sources form networks in order to exchange information. As a direct consequence, the field of Graph Signal Processing has been thriving and evolving. Its aim: process and make sense of all the surrounding data deluge.In this context, the main goal of this thesis is developing methods and algorithms capable of using data streams, in a distributed fashion, in order to infer the underlying networks that link these streams. Then, these estimated network topologies can be used with tools developed for Graph Signal Processing in order to process and analyze data supported by graphs. After a brief introduction followed by motivating examples, we first develop and propose an online, distributed and adaptive algorithm for graph topology inference for data streams which are linearly dependent. An analysis of the method ensues, in order to establish relations between performance and the input parameters of the algorithm. We then run a set of experiments in order to validate the analysis, as well as compare its performance with that of another proposed method of the literature.The next contribution is in the shape of an algorithm endowed with the same online, distributed and adaptive capacities, but adapted to inferring links between data that interact non-linearly. As such, we propose a simple yet effective additive model which makes use of the reproducing kernel machinery in order to model said nonlinearities. The results if its analysis are convincing, while experiments ran on biomedical data yield estimated networks which exhibit behavior predicted by medical literature.Finally, a third algorithm proposition is made, which aims to improve the nonlinear model by allowing it to escape the constraints induced by additivity. As such, the newly proposed model is as general as possible, and makes use of a natural and intuitive manner of imposing link sparsity, based on the concept of partial derivatives. We analyze this proposed algorithm as well, in order to establish stability conditions and relations between its parameters and its performance. A set of experiments are ran, showcasing how the general model is able to better capture nonlinear links in the data, while the estimated networks behave coherently with previous estimates

Style APA, Harvard, Vancouver, ISO itp.

30

Renoust, Benjamin. "Analysis and Visualisation of Edge Entanglement in Multiplex Networks". Phd thesis, Université Sciences et Technologies - Bordeaux I, 2013. http://tel.archives-ouvertes.fr/tel-00942358.

Pełny tekst źródła

Streszczenie:

When it comes to comprehension of complex phenomena, humans need to understand what interactions lie within them.These interactions are often captured with complex networks. However, the interaction pluralism is often shallowed by traditional network models. We propose a new way to look at these phenomena through the lens of multiplex networks, in which catalysts are drivers of the interaction through substrates. To study the entanglement of a multiplex network is to study how edges intertwine, in other words, how catalysts interact. Our entanglement analysis results in a full set of new objects which completes traditional network approaches: the entanglement homogeneity and intensity of the multiplex network, and the catalyst interaction network, with for each catalyst, an entanglement index. These objects are very suitable for embedment in a visual analytics framework, to enable comprehension of a complex structure. We thus propose of visual setting with coordinated multiple views. We take advantage of mental mapping and visual linking to present simultaneous information of a multiplex network at three different levels of abstraction. We complete brushing and linking with a leapfrog interaction that mimics the back-and-forth process involved in users' comprehension. The method is validated and enriched through multiple applications including assessing group cohesion in document collections, and identification of particular associations in social networks.

Style APA, Harvard, Vancouver, ISO itp.

31

Turco, Antonio. "Use of carbon nanotubes for novel approaches towards spinal network repairing". Doctoral thesis, Università degli studi di Trieste, 2013. http://hdl.handle.net/10077/8663.

Pełny tekst źródła

Streszczenie:

2011/2012
Nanotechnology underwent a very rapid development in the last decades, thanks to the invention of different techniques that allow reaching the nanoscale. The great interest in this area arises from the variety of possible applications in different fields, such as electronics, where the miniaturization of components is a key factor, but also medicine. The creation of smart systems able to carry out a specific task in the body in a controlled way, either in diagnosis or therapy or tissue engineering, is the ultimate goal of a newborn area of research, called nanomedicine. In fact, to reach such an outstanding objective, a nanometer‐sized material is needed and carbon nanotubes (CNTs) are among the most promising candidates. The aim of this thesis was to study this opportunity and, in particular, the possible application of carbon nanotubes for spinal network repairing. After a review of the main features of neuronal network systems and the most common techniques to study their functionality, possible applications of nanotechnology for nanomedicine purposes are considered, focusing the attention on CNTs as neuronal interface in nerve tissue engineering. The work can be divided into two big parts. In the first part the impact of carbon nanotubes on various neuronal systems was studied. Different form of carbonaceous materials (carbon nanotubes, nanohorns and graphene) were deposited in a homogeneous way on a glass surface playing with organic functionalization and different deposition techniques. Hippocampal neuronal cells were grown on their surface to better understand how morphology and conductivity of the material could influence the activity of the neuronal network evidencing how both these characteristics could affect the electrophysiological properties of neurons. Then, also spinal neurons were grown on carbon nanotubes network deposited on a glass substrate to evaluate, for the first time, the impact of carbon nanotubes on this kind of cells. The tight interaction between these two materials appeared to cause a faster maturation of the spinal neurons with respect II to the control grown on a glass substrate. The long-term impact on a complex tissue (spinal cord slice) grown on carbon nanotubes carpet was also studied. The intimate interaction between the two materials observed by TEM and SEM analysis caused an increase in dimensions and number of neuronal fibers that comes out from the body of a spinal cord slice. An increase in electrophysiological activity of all neuronal network of the slice was also reported. In the second part of the work different conductive biocompatible nanocomposite materials based on carbon nanotubes and “artificial” polymers (such as Nafion, PVA, PET, PEI, PDMS and PANI) were investigated. The idea is to test these materials as neuronal prosthesis to repair spinal cord damage. All the prepared scaffolds showed CNTs on the surface favoring CNTs-neurons interaction. To address this aim different techniques and different organic functionalizations of CNTs were utilized to control supramolecular interactions between the nanomaterial and polymers orienting the deposition of the CNTs and preventing their aggregation. After that, an innovative method to study the possible ability of this nanocomposite materials to transmit a neuronal signal between two portions of spinal cord was designed. Functionalization of gold surfaces with thiolated carbon nanotubes have been conducted in order to develop suitable devices for neuronal stimulation and consequent spinal cord lesions repairing. In particular thiol groups were introduced on the graphitic surface of carbon nanotubes by means of covalent functionalization. First of all, the interaction of CNTs with gold nanoparticles has been evaluated, then a gold surface has been coated by means of contact printing technique with a homogeneous film of CNTs. This hybrid material could be useful to produce innovative electrodes for neuronal stimulation
XXV Ciclo
1985

Style APA, Harvard, Vancouver, ISO itp.

32

Combe, David. "Détection de communautés dans les réseaux d'information utilisant liens et attributs". Phd thesis, Université Jean Monnet - Saint-Etienne, 2013. http://tel.archives-ouvertes.fr/tel-01056985.

Pełny tekst źródła

Streszczenie:

Alors que les réseaux sociaux s'attachent à représenter des entités et les relations existant entre elles, les réseaux d'information intègrent également des attributs décrivant ces entités ; ce qui conduit à revisiter les méthodes d'analyse et de fouille de ces réseaux. Dans ces travaux, nous proposons des méthodes de classification des entités du réseau d'information qui exploitent d'une part les relations entre celles-ci et d'autre part les attributs les caractérisant. Nous nous penchons sur le cas des réseaux à vecteurs d'attributs, où les entités du réseau sont décrites par des vecteurs numériques. Ainsi nous proposons des approches basées sur des techniques reconnues pour chaque type d'information, faisant appel notamment à l'inertie pour la classification automatique et à la modularité de Newman et Girvan pour la détection de communautés. Nous évaluons nos propositions sur des réseaux issus de données bibliographiques, faisant usage en particulier d'information textuelle. Nous évaluons également nos approches face à diverses évolutions du réseau, notamment au regard d'une détérioration des informations des liens et des attributs, et nous caractérisons la robustesse de nos méthodes à celle-ci

Style APA, Harvard, Vancouver, ISO itp.

33

BOSCHI, Alex. "Studio dei meccanismi di trasporto di carica in film sottili a base di materiali correlati al grafene (GRM)". Doctoral thesis, Università degli studi di Modena e Reggio Emilia, 2021. http://hdl.handle.net/11380/1244690.

Pełny tekst źródła

Streszczenie:

L’avvento di tecniche economiche per produrre fogli di materiali con spessore mono-atomico, come il grafene [1], ha aperto la possibilità di progettare materiali nanostrutturati con proprietà chimico/fisiche pre-programmabili. I sistemi tecnologicamente rilevanti a base di materiali correlati al grafene (GRM) sono principalmente networks di microfogli bidimensionali distribuiti in modo casuale e altamente difettosi [2]. Sebbene il trasporto di carica sia stato ampiamente trattato per fogli bidimensionali isolati [3], uno studio esaustivo che correli le proprietà elettriche dei GRM networks con la complessità della struttura e della morfologia del materiale è ancora mancante in letteratura. Lo scopo di questo lavoro è indagare il trasporto di carica (CT) in film sottili di GRM, con un certo grado di disordine strutturale. In particolare, abbiamo studiato i meccanismi di CT che si verificano all'interfaccia tra due fogli – spesso i meccanismi d’interfaccia sono considerati un fattore limitante – così come il ruolo della complessità geometrica del network nella conducibilità elettrica complessiva delle strutture di microfogli. Abbiamo utilizzato fogli monoatomici di grafene ossido (GO) come materiale bidimensionale modello. Il GO consiste infatti in un reticolo di grafene conduttivo che include gruppo funzionali dell’ossigeno sia sul piano basale che ai bordi del foglio. I microfogli isolanti di GO vengono depositati su substrati di ossido di silicio e sono ridotti termicamente per ripristinare alcune delle proprietà del grafene, tra cui la conducibilità elettrica. Oltre al GO ridotto, abbiamo impiegato un GRM composto da fogli multistrato di grafene (parzialmente ossidato). Tale materiale è il GO esfoliato elettrochimicamente (eGO). Abbiamo quindi sfruttato diversi metodi di deposizione: i) spin-coating, ii) spray-coating e iii) filtrazione assistita da vuoto per fabbricare film sottili di fogli di GRM parzialmente impilati e con dimensioni laterali macroscopiche. Le proprietà chimiche e morfologiche dei film sono state caratterizzate attraverso misurazioni di spettroscopia fotoelettronica a raggi X (XPS), microscopia a forza atomica (AFM) e diffrazione di raggi X (XRD). I meccanismi di trasporto sono stati studiati invece con misure di resistività elettrica (ρ) in funzione della temperatura, da valori ambientali fino a 5 K. Possibili ambiguità sull'analisi quantitativa di ρ(T) sono state risolte utilizzando un metodo autoconsistente basato sull'energia di attivazione ridotta [5], ovvero la derivata logaritmica della resistività rispetto alla temperatura: W(T)=-(d ln⁡ρ)⁄(d ln⁡T ). Questa trasformazione matematica ha permesso di analizzare il dataset ρ(T) con funzioni lineari. Abbiamo correlato i parametri caratteristici di trasporto con il grado di ordine dei nostri campioni e chiarito il ruolo dell'impacchettamento verticale dei fogli, ovvero dell'interazione π-π tra cluster aromatici sovrapposti, nel CT del film. Abbiamo inoltre evidenziato le differenze di CT tra i film basati su GO ridotto e quelli su eGO. Il lavoro proposto pone le basi per lo sviluppo di nuovi modelli e protocolli riguardanti i meccanismi di CT in sistemi realistici di GRM, come inchiostri e compositi polimerici. [1] Ferrari, A. C. et al. Nanoscale 7, 4598-4810, (2015). [2] Palermo V., Chem. Comm. 49, 28, 2848-2857 (2013); Kelly A. et al, Science 356, 6333 (2017). [3] Eda G. et al, J. Physics. Chem.C 113, 15768 (2009); Kaiser a. et al, Nano Letters 9, 1787 (2009); Joung D. and Khondaker S., Phys. Rev. B 86, 235423 (2012). [4] Xia Z. et al, J. Physics. Chem.C 123, 15122 (2019). [5] Zabrodskii A. G., Philos. Mag. B 81, 1131 (2001).
The development of cheap techniques to produce large sheets of monoatomic thick materials, such as graphene [1], opened new avenues to design nanostructured materials with pre-programmed chemical and physical properties. Most of the technologically relevant graphene-related materials (GRMs) systems are networks composed of randomly distributed and highly defective 2D microsheets [2]. While the charge transport has been extensively studied in single nanosheets [3], a comprehensive study that correlates the electrical properties of networks composed of purely 2D graphene-based materials with the complexity of the material structure and morphology is still missing. The aim of this work is to investigate charge transport (CT) in GRMs films, going towards structures with increasing disorder. In particular we investigated the CT mechanisms occurring at the sheet-to-sheet interface – typically the interfacial mechanisms are considered as bottlenecks – as well as the role of the geometrical complexity of the network in the overall electrical conductivity of the nanosheets assemblies. As prototypical 2D material we used single monolayer sheets of graphene oxide (GO), which consists of a conductive graphene lattice including oxygen functionalities/ defects both on the basal plane and at the edges of the sheet. Electrical insulating GO sheets are deposited on silicon oxide substrates and thermally reduced restoring partially the conductive properties of the 2D sheets. In addition to reduced GO, we employed a GRM made of multiple staked sheets of (partially oxidised) graphene bilayers: electrochemical exfoliated GO (eGO) [4].We exploited different deposition methods: i) spin-coating, ii) spray-coating and iii) vacuum-assisted filtration to fabricate macroscopic GRMs thin films with sheets partially stacked. Chemical and morphological properties of the films were characterized by X-ray Photoelectron Spectroscopy (XPS), Atomic Force Microscopy (AFM) and X-ray Diffraction (XRD) measurements. We investigated transport mechanisms measuring the temperature-dependence of the electrical resistivity (ρ) from room temperature down to 5 K. Possible ambiguities on the quantitative analysis of ρ(T) were solved by using a robust self-consistent method based on the reduced activation energy [5], i.e. the logarithmic derivative of resistivity versus temperature: W(T)=-(d ln⁡ρ)⁄(d ln⁡T ). This mathematical transformation allowed to analyse ρ(T) dataset with linear functions. We correlated the transport characteristic parameters with the degree of order of our samples and elucidate the role of the sheets vertical stacking, that is of the π-π interaction between overlapped aromatic clusters, in the CT in the film. We also highlighted the differences in CT between reduced GO based films and eGO ones. The presented work could pave the way to develop new models and protocols to access the CT mechanisms in realistic GRMs, such as inks and polymer composites. [1] Ferrari, A. C. et al. Nanoscale 7, 4598-4810, (2015). [2] Palermo V., Chem. Comm. 49, 28, 2848-2857 (2013); Kelly A. et al, Science 356, 6333 (2017). [3] Eda G. et al, J. Physics. Chem.C 113, 15768 (2009); Kaiser a. et al, Nano Letters 9, 1787 (2009); Joung D. and Khondaker S., Phys. Rev. B 86, 235423 (2012). [4] Xia Z. et al, J. Physics. Chem.C 123, 15122 (2019). [5] Zabrodskii A. G., Philos. Mag. B 81, 1131 (2001).

Style APA, Harvard, Vancouver, ISO itp.

34

Wauquier, Pauline. "Task driven representation learning". Thesis, Lille 3, 2017. http://www.theses.fr/2017LIL30005/document.

Pełny tekst źródła

Streszczenie:

De nombreux algorithmes d'Apprentissage automatique ont été proposés afin de résoudre les différentes tâches pouvant être extraites des problèmes de prédiction issus d'un contexte réel. Pour résoudre les différentes tâches pouvant être extraites, la plupart des algorithmes d'Apprentissage automatique se basent d'une manière ou d'une autre sur des relations liant les instances. Les relations entre paires d'instances peuvent être définies en calculant une distance entre les représentations vectorielles des instances. En se basant sur la représentation vectorielle des données, aucune des distances parmi celles communément utilisées n'est assurée d'être représentative de la tâche à résoudre. Dans ce document, nous étudions l'intérêt d'adapter la représentation vectorielle des données à la distance utilisée pour une meilleure résolution de la tâche. Nous nous concentrons plus précisément sur l'algorithme existant résolvant une tâche de classification en se basant sur un graphe. Nous décrivons d'abord un algorithme apprenant une projection des données dans un espace de représentation permettant une résolution, basée sur un graphe, optimale de la classification. En projetant les données dans un espace de représentation dans lequel une distance préalablement définie est représentative de la tâche, nous pouvons surpasser la représentation vectorielle des données lors de la résolution de la tâche. Une analyse théorique de l'algorithme décrit est développée afin de définir les conditions assurant une classification optimale. Un ensemble d'expériences nous permet finalement d'évaluer l'intérêt de l'approche introduite et de nuancer l'analyse théorique
Machine learning proposes numerous algorithms to solve the different tasks that can be extracted from real world prediction problems. To solve the different concerned tasks, most Machine learning algorithms somehow rely on relationships between instances. Pairwise instances relationships can be obtained by computing a distance between the vectorial representations of the instances. Considering the available vectorial representation of the data, none of the commonly used distances is ensured to be representative of the task that aims at being solved. In this work, we investigate the gain of tuning the vectorial representation of the data to the distance to more optimally solve the task. We more particularly focus on an existing graph-based algorithm for classification task. An algorithm to learn a mapping of the data in a representation space which allows an optimal graph-based classification is first introduced. By projecting the data in a representation space in which the predefined distance is representative of the task, we aim at outperforming the initial vectorial representation of the data when solving the task. A theoretical analysis of the introduced algorithm is performed to define the conditions ensuring an optimal classification. A set of empirical experiments allows us to evaluate the gain of the introduced approach and to temper the theoretical analysis

Style APA, Harvard, Vancouver, ISO itp.

35

Schiller, Benjamin. "Graph-based Analysis of Dynamic Systems". Doctoral thesis, Saechsische Landesbibliothek- Staats- und Universitaetsbibliothek Dresden, 2017. http://nbn-resolving.de/urn:nbn:de:bsz:14-qucosa-230611.

Pełny tekst źródła

Streszczenie:

The analysis of dynamic systems provides insights into their time-dependent characteristics. This enables us to monitor, evaluate, and improve systems from various areas. They are often represented as graphs that model the system's components and their relations. The analysis of the resulting dynamic graphs yields great insights into the system's underlying structure, its characteristics, as well as properties of single components. The interpretation of these results can help us understand how a system works and how parameters influence its performance. This knowledge supports the design of new systems and the improvement of existing ones. The main issue in this scenario is the performance of analyzing the dynamic graph to obtain relevant properties. While various approaches have been developed to analyze dynamic graphs, it is not always clear which one performs best for the analysis of a specific graph. The runtime also depends on many other factors, including the size and topology of the graph, the frequency of changes, and the data structures used to represent the graph in memory. While the benefits and drawbacks of many data structures are well-known, their runtime is hard to predict when used for the representation of dynamic graphs. Hence, tools are required to benchmark and compare different algorithms for the computation of graph properties and data structures for the representation of dynamic graphs in memory. Based on deeper insights into their performance, new algorithms can be developed and efficient data structures can be selected. In this thesis, we present four contributions to tackle these problems: A benchmarking framework for dynamic graph analysis, novel algorithms for the efficient analysis of dynamic graphs, an approach for the parallelization of dynamic graph analysis, and a novel paradigm to select and adapt graph data structures. In addition, we present three use cases from the areas of social, computer, and biological networks to illustrate the great insights provided by their graph-based analysis. We present a new benchmarking framework for the analysis of dynamic graphs, the Dynamic Network Analyzer (DNA). It provides tools to benchmark and compare different algorithms for the analysis of dynamic graphs as well as the data structures used to represent them in memory. DNA supports the development of new algorithms and the automatic verification of their results. Its visualization component provides different ways to represent dynamic graphs and the results of their analysis. We introduce three new stream-based algorithms for the analysis of dynamic graphs. We evaluate their performance on synthetic as well as real-world dynamic graphs and compare their runtimes to snapshot-based algorithms. Our results show great performance gains for all three algorithms. The new stream-based algorithm StreaM_k, which counts the frequencies of k-vertex motifs, achieves speedups up to 19,043 x for synthetic and 2882 x for real-world datasets. We present a novel approach for the distributed processing of dynamic graphs, called parallel Dynamic Graph Analysis (pDNA). To analyze a dynamic graph, the work is distributed by a partitioner that creates subgraphs and assigns them to workers. They compute the properties of their respective subgraph using standard algorithms. Their results are used by the collator component to merge them to the properties of the original graph. We evaluate the performance of pDNA for the computation of five graph properties on two real-world dynamic graphs with up to 32 workers. Our approach achieves great speedups, especially for the analysis of complex graph measures. We introduce two novel approaches for the selection of efficient graph data structures. The compile-time approach estimates the workload of an analysis after an initial profiling phase and recommends efficient data structures based on benchmarking results. It achieves speedups of up to 5.4 x over baseline data structure configurations for the analysis of real-word dynamic graphs. The run-time approach monitors the workload during analysis and exchanges the graph representation if it finds a configuration that promises to be more efficient for the current workload. Compared to baseline configurations, it achieves speedups up to 7.3 x for the analysis of a synthetic workload. Our contributions provide novel approaches for the efficient analysis of dynamic graphs and tools to further investigate the trade-offs between different factors that influence the performance.

Style APA, Harvard, Vancouver, ISO itp.

36

Vieira, Milreu Paulo. "Enumerating functional substructures of genome-scale metabolic networks : stories, precursors and organisations". Phd thesis, Université Claude Bernard - Lyon I, 2012. http://tel.archives-ouvertes.fr/tel-00850704.

Pełny tekst źródła

Streszczenie:

In this thesis, we presented three different methods for enumerating special subnetworks containedin a metabolic network: metabolic stories, minimal precursor sets and chemical organisations. Foreach of the three methods, we gave theoretical results, and for the two first ones, we further providedan illustration on how to apply them in order to study the metabolic behaviour of living organisms.Metabolic stories are defined as maximal directed acyclic graphs whose sets of sources and targets arerestricted to a subset of the nodes. The initial motivation of this definition was to analyse metabolomicsexperimental data, but the method was also explored in a different context. Metabolic precursor setsare minimal sets of nutrients that are able to produce metabolites of interest. We present threedifferent methods for enumerating minimal precursor sets and we illustrate the application in a studyof the metabolic exchanges in a symbiotic system. Chemical organisations are sets of metabolites thatare simultaneously closed and self-maintaining, which captures some stability feature in the

Style APA, Harvard, Vancouver, ISO itp.

37

Tosatto, Silvia. "A teledetection system based on surface plasmon resonance sensors for environmental pollutants". Doctoral thesis, Università degli studi di Padova, 2013. http://hdl.handle.net/11577/3422980.

Pełny tekst źródła

Streszczenie:

Surface Plasmon Resonance (SPR) sensors are thin-film refractometers that measure changes in the refractive index that take place at the surface of a metal film supporting a surface plasmon (Homola, 2008). The research activity consisted on a study and realization of innovative Surface Plasmon Resonance based sensors and of their applications, in particular for the environment monitoring. In order to investigate the possibility of creating innovative SPR sensors, computer simulations of optical structures supporting surface plasmon polaritons and comprising peculiar materials were implemented. In particular SPR sensors based on metals showing Inverted Surface Plasmon Resonance and a single layer of Graphene (SGL) were simulated. An article has been submitted on this study. Surface Plasmon Resonance sensor prototypes were also realized in laboratory. By means of a first SPR sensor with dynamic setup, measurements with Helium and Pentane gases were made, then an SPR sensor with a static setup was assembled, refined and as a following improvement in the optical components an adaptive mirror also introduced, which was able to counteract, if present, a lens defocus, spherical aberration, coma and astigmatism . Moreover the elements of a feasible wireless sensor network based on Surface Plasmon Resonance sensors, the possible data that can be inferred from a SPR sensor and a possible data handling strategy were determined along with an evaluation of the sensor module energy requirements. Then two small wireless sensor network test setups were implemented in laboratory, each one composed of two SPR sensors controlled by a computer or a microcontroller, ZigBee antennas and a main computer for the data reception and diffusion into Internet. The feasibility of a Wireless Sensor Network based on SPR sensors, conceived for environmental pollutants detection and with broadband internet connection, has been demonstrated and two different laboratory setups for a test Wireless Sensor Network based on Surface Plasmon Resonance Sensors were realized.
I sensori basati sulla plasmonica di superficie (SPR) sono rifrattometri, basati sulla fisica dei film sottili, che misurano le variazioni di indice di rifrazione che avvengono sulla superficie di uno strato metallico supportante un plasmone di superficie (Homola , 2008). L’ attività di ricerca si è sviluppata come uno studio ed implementazione di sensori SPR innovativi e delle loro possibili applicazioni, in particolare per la rilevazione di sostanze inquinanti. Sono stati simulati in ambiente Matlab sensori SPR innovativi, comprendenti materiali peculiari. Tali sensori comprendono in particolare metalli mostranti una Risonanza Plasmonica Invertita (ISPR), oltre che un singolo strato di Graphene (SGL). Un articolo è stato sottomesso su tali studi. Sono stati inoltre implementati in laboratorio vari prototipi di sensori basati sulla plasmonica di superficie. Innanzi tutto sono state effettuate delle misure di riflettività con i gas elio e pentano mediante un sensore basato sulla plasmonica di superficie e con un setup dinamico, quindi è stato creato un nuovo sensore utilizzante un un sistema ottico di tipo statico ed un prisma cilindrico ad alto indice di rifrazione, migliorando il più possibile la qualità del fascio ottico e filtrandolo opportunamente. Altresì è stato innovativamente utilizzato nel setup ottico uno specchio deformabile, in grado di correggere le aberrazioni ottiche presenti nello stesso. Oltre a ciò sono stati individuati i possibili elementi di una rete wireless di sensori SPR progettata per il monitoraggio ambientale, i dati ottenibili da un sensore SPR, una strategia per la diffusione dei dati ed una stima dei consumi energetici. Infine sono stati implementati in laboratorio due diversi setup di prova per una piccola rete wireless di sensori, formata da due sensori SPR controllati da un computer o da un microcontrollore, antenne ZigBee e da un computer principale per la ricezione, analisi e diffusione in Internet dei dati. L’ attività di ricerca presentata in questa tesi ha quindi dimostrato la possibilità della creazione di una rete wireless di sensori basati sulla Plasmonica di Superficie e con accesso Internet a banda larga, finalizzata alla rilevazione di sostanze inquinanti in ambiente acquoso, e due piccole versioni di prova della stessa sono state implementate in laboratorio.

Style APA, Harvard, Vancouver, ISO itp.

38

Candell, Richard. "Performance Estimation, Testing, and Control of Cyber-Physical Systems Employing Non-ideal Communications Networks". Thesis, Bourgogne Franche-Comté, 2020. http://www.theses.fr/2020UBFCK017.

Pełny tekst źródła

Streszczenie:

La technologie sans fil est un catalyseur clé des promesses de l’industrie 4.0 (fabrication intelligente). En tant que telle, la technologie sans fil sera adoptée comme mode de communication principal au sein de l’usine en général et dans les unités de production en particulier. La communication des unités de production en usine a des exigences particulières en matière de latence, de fiabilité, d’échelle et de sécurité qui doivent d’abord être satisfaites par la technologie de communication sans fil utilisée. Le sans fil est considéré comme une forme de communication non idéale dans la mesure où par rapport aux communications câblées, il est considéré comme moins fiable (avec perte) et moins sécurisé. Ces dégradations possibles entraînent un retard et une perte de données dans un système d’automatisation industrielle où le déterminisme, la sécurité et la sûreté sont considérés comme primordiaux. Cette thèse étudie les exigences d’une communication sans fil dans les unités de production et l’applicabilité de la technologie sans fil existante dans ce domaine. Elle présente une modélisation SysML de l’architecture du système et des flux de données. Elle fournit une méthode d’utilisation des bases de données de type graphe pour l’organisation et l’analyse des données de performance collectées à partir d’un environnement de test. Enfin, la thèse décrit une approche utilisant l’apprentissage automatique pour l’évaluation des performances d’un système d’objets connectés dans le domaine de fabrication
Wireless technology is a key enabler of the promises of Industry 4.0 (Smart Manufacturing). As such, wireless technology will be adopted as a principal mode of communication within the factory beginning with the factory enterprise and eventually being adopted for use within the factory workcell. Factory workcell communication has particular requirements on latency, reliability, scale, and security that must first be met by the wireless communication technology used. Wireless is considered a non-ideal form of communication in that when compared to its wired counterparts, it is considered less reliable (lossy) and less secure. These possible impairments lead to delay and loss of data in industrial automation system where determinism, security, and safety is considered paramount. This thesis investigates the wireless requirements of the factory workcell and applicability of existing wireless technology, it presents a modeling approach to discovery of architecture and data flows using SysML, it provides a method for the use of graph databases to the organization and analysis of performance data collected from a testbed environment, and finally provides an approach to using machine learning in the evaluation of cyberphysical system performance

Style APA, Harvard, Vancouver, ISO itp.

39

Tackx, Raphaël. "Analyse de la structure communautaire des réseaux bipartis". Electronic Thesis or Diss., Sorbonne université, 2018. https://accesdistant.sorbonne-universite.fr/login?url=https://theses-intra.sorbonne-universite.fr/2018SORUS550.pdf.

Pełny tekst źródła

Streszczenie:

Il existe dans le monde réel un nombre important de réseaux qui apparaissent naturellement, on les retrouve un peu partout, dans de nombreuses disciplines, par exemple en informatique avec les réseaux de routeurs, les réseaux de satellites, les réseaux de pages Web, en biologie avec les réseaux des neurones, en écologie avec les réseaux d’interactions biologiques, en linguistiques avec les réseaux de synonymes, en droit avec les réseaux de décisions juridiques, en économie avec les réseaux interbancaires, en sciences humaines avec les réseaux sociaux. De manière générale, un réseau reflète les interactions entre les nombreuses entités d’un système. Ces interactions peuvent être de différentes natures, un lien social ou un lien d’amitié dans un réseau social constitué de personnes, un câble dans un réseau de routeurs, une réaction chimique dans un réseau biologique de protéines, un hyperlien dans un réseau de pages Web, etc. Plus encore, la rapide démocratisation du numérique dans nos sociétés, avec Internet notamment, a pour conséquence de produire de nouveaux systèmes qui peuvent être représentés sous forme de réseaux. Finalement, tous ces réseaux présentent des particularités bien spécifiques : ils sont issus de contextes pratiques, ils sont le plus souvent de grande taille (on retrouve quelques fois des réseaux constitués de plusieurs milliards de nœuds et de liens, contenant donc une grande quantité d’information), ils présentent des propriétés statistiques communes. À cet égard, ils sont regroupés sous l’appellation de réseaux réels, graphes de terrain ou encore réseaux complexes. Aujourd'hui, la science des réseaux est un domaine de recherche à part entière dont l’enjeu principal est de parvenir à décrire et modéliser ces réseaux avec précision afin de révéler leurs caractéristiques générales et de mieux comprendre leurs mécanismes. La plupart des travaux dans ce domaine utilisent le formalisme des graphes qui fournit un ensemble d’outils mathématiques particulièrement adaptés à l’analyse topologique et structurelle des réseaux. Il existe de nombreuses applications dans ce domaine, par exemple des applications concernant la propagation d’épidémie ou de virus informatique, la fragilité du réseau en cas de panne, sa résilience en cas d’attaque, l’étude de la dynamique pour prédire l’apparition de nouveaux liens, la recommandation, etc. L’un des problèmes complexes actuels, qui a beaucoup d’applications, est l’identification de la structure communautaire. La grande majorité des réseaux réels sont caractérisés par des niveaux d’organisation dans leur structure mésoscopique. Du fait de la faible densité globale des réseaux réels couplée à la forte densité locale, on observe la présence de groupes de nœuds fortement liés entre eux et plus faiblement liés avec le reste du réseau, que l’on appelle communautés. Ces structures ont également du sens dans le réseau lui-même, par exemple les communautés d’un réseau social peuvent correspondre à des groupes sociaux (amis, familles, etc.), les communautés d’un réseau de protéines peuvent traduire des réponses fonctionnelles, elles peuvent correspondre à des sujets similaires dans un réseau de pages Web, pour donner quelques exemples [...]
In the real world, numerous networks appear naturally, they are everywhere, in many disciplines, for example in computer science with router networks, satellite networks, webpage networks, in biology with neural networks, in ecology with biological interaction networks, in linguistic with synonym networks, in law with legal decision networks, in economy with interbank networks, in social sciences and humanities with social networks. Generally, a network reflects the interactions between many entities of a system. These interactions have different sources, a social link or a friendship link in a social network, a cable in a router network, a chemical reaction in a protein-protein interaction network, a hyperlink in a webpage network. Furthermore, the rapid democratization of digital technology in our societies, with internet in particular, leads to create new systems which can be seen as networks. Finally, all these networks depict very specific features : they come from pratical contexts, most of the time they are big (they may be comprised of several billion of nodes and links, containing a large amount of information), they share statistical properties. In this regard, they are called real-world networks or complex networks. Nowaday, network science is a research area in its own right focusing on describing and modeling these networks in order to reveal their main features and improve our understanding of their mecanisms. Most of the works in this area use graphs formalism which provides a set of mathematical tools well suited for analyzing the topology of these networks. It exists many applications, for instance applications in spread of epidemy or computer viruses, weakness of networks in case of a breakdown, attack resilience, study for link prediction, recommandation, etc. One of the major issue is the identification of community structure. The large majority of real-world networks depicts several levels of organization in their structure. Because of there is a weak global density coupled with a strong local density, we observe that nodes are usually organized into groups, called communities, which are more internally connected than they are to the rest of the network. Moreover, these structures have a meaning in the network itself, for example communities of a social network may correspond to social groups (friends, families, etc.), communities of a protein-protein network may translate fonctions of a cell, communities may be also related to similar subjects in a webpage network [...]

Style APA, Harvard, Vancouver, ISO itp.

40

Costes, Benoît. "Vers la construction d'un référentiel géographique ancien : un modèle de graphe agrégé pour intégrer, qualifier et analyser des réseaux géohistoriques". Thesis, Paris Est, 2016. http://www.theses.fr/2016PESC1032/document.

Pełny tekst źródła

Streszczenie:

Les historiens et archéologues ont efficacement mis à profit les travaux réalisés dans le domaine des SIG pour répondre à leurs propres problématiques. Pour l'historien, un Système d’Information Géographique est avant tout un outil de compréhension des phénomènes sociaux.De nombreuses sources géohistoriques sont aujourd'hui mises à la disposition des chercheurs: plans anciens, bottins, etc. Le croisement de ces sources d'informations diverses et hétérogènes soulève de nombreuses questions autour des dynamiques urbaines.Mais les données géohistoriques sont par nature imparfaites, et pour pouvoir être exploitées, elles doivent être spatialisées et qualifiées.L'objectif de cette thèse est d'apporter une solution à ce verrou par la production de données anciennes de référence. En nous focalisant sur le réseau des rues de Paris entre la fin du XVIIIe et la fin du XIXe siècles, nous proposons plus précisément un modèle multi-représentations de données agrégées permettant, par confrontation d'observations homologues dans le temps, de créer de nouvelles connaissances sur les imperfections des données utilisées et de les corriger. Nous terminons par tester le rôle de référentiel géohistorique des données précédemment qualifiées et enrichies en spatialisant et intégrant dans le modèle de nouvelles données géohistoriques de types variés (sociales et spatiales), en proposant de nouvelles approches d'appariement et de géocodage
The increasing availability of geohistorical data, particularly through the development of collaborative projects is a first step towards the design of a representation of space and its changes over time in order to study its evolution, whether social, administrative or topographical.Geohistorical data extracted from various and heterogeneous sources are highly inaccurate, uncertain or inexact according to the existing terminology. Before being processed, such data should be qualified and spatialized.In this thesis, we propose a solution to this issue by producing reference data. In particular, we focus on Paris historical street networks and its evolution between the end of the XVIIIth and the end of the XIXth centuries.Our proposal is based on a merged structure of multiple representations of data capable of modelling spatial networks at different times, providing tools such as pattern detection in order to criticize, qualify and eventually correct data and sources without using ground truth data but the comparison of data with each other through the merging process.Then, we use the produced reference data to spatialize and integrate other geohistorical data such as social data, by proposing new approaches of data matching and geocoding

Style APA, Harvard, Vancouver, ISO itp.

41

Damaj, Samer. "SWH, application "Small-world" à la génération des réseaux d'interconnexion pour les architectures massivement parallèles". Brest, 2011. http://www.theses.fr/2011BRES2076.

Pełny tekst źródła

Streszczenie:

Les possibilités d'intégration sont de plus en plus importantes et le nombre de transistors disponibles sur un composant n'est plus aussi critique que dans le passé. Cependant les structures de traitement du type monoprocesseur ne bénéficient plus de cette augmentation de ressources et on voit apparaître des composants du type MP Soc (multiprocessors on Chip) qui intègre plusieurs coeurs de processeur sur une même puce. Le problème consiste donc à faire passer à l'échelle des architectures systèmes du type MIMD (Multiple Instruction stream Multiple Data stream). Si on considère des coeurs de processeur complexe alors le problème reste actuellement assez maîtrisable car leur nombre reste limité (2 voit 4 processeurs). Par contre l'utilisation de noeuds de traitement de plus petite taille autorise l'intégration de dizaines voire de centaines de processeurs sur un même circuit. Les besoins en puissance de calcul ne se satisfont plus des limites rencontrées sur des processeurs complexes. La généralisation des composants du type MPSoC qui intègrent un nombre toujours plus important de processeurs sur une puce. Cependant, interconnecter un grand nombre de processeurs sur une puce est un délicat. Essentiellement, les structures adaptées aux contraintes de la puce sont trop peu générales, et les schémas de l’interconnexion plus globale engendrent des coûts de fabrication trop élevés. Ce problème se décompose en deux sous problèmes intimement liés qui sont la recherche d'une topologie permettant d'atteindre un niveau de connectivité acceptable et le coût matériel de sa réalisation. C'est dans ce dernier cas qui s'inscrit le contexte scientifique de ce mémoire. Les concepteurs se basent sur des différentes architectures pour réaliser des puces. Les architectures à mémoire partagée où il y a une saturation d’accès mémoire avec un grand nombre de PEs (Processeurs Élémentaires) comme par exemple TILERA Corporation. Les architectures à mémoire distribué où il se manifeste le problème des programmes fortement communicants. Ce type d'architecture est structuré autour d'un Noc (Network-on-Chip). De nombreuses études ont déjà été réalisées sur la problématique des architectures et de leurs réalisations avec pour motivations de trouver une manière élégante de supporter des Pes qui est de plus en plus nombreux et rapide en limitant le nombre de fils et pour obtenir des réseaux réguliers où le coût et le temps de développement du layout sont très court. Pour cette raison, nous sommes à la recherche d'une nouvelle plate-forme homogène avec des multicoeurs réguliers et massivement parallèles. C'est dans ce dernier cas qui s'inscrit le contexte industriel de ce mémoire de thèse. Pour des raisons techniques, il faut rendre les circuits homogènes, réguliers et denses parce que le coût et le temps de développement du layout d'un circuit non régulier sont très longs. Pour cette raison, il faut s'appuyer sur des architectures massivement parallèles. Les processeurs doivent communiquer entre eux, ils sont donc connectés via un réseau qui implémente une topologie. Pour une topologie donnée, on dispose de métriques grossières telles que le nombre de noeuds (le nombre de sommets dans un graphe donné), le degré (le nombre de liens entrants et sortants pour chaque sommet dans ce graphe donné) et le diamètre (la distance maximale entre deux noeuds ou sommets quelconques dans le graphe). Dans ce cas, comme on veut un circuit régulier, où les CPUs sont organisés en grille 2D pour les techniques actuels et éventuellement 3D pour les futurs techniques alors il faut obtenir le coût silicium pour mapper la topologie choisie sur cette grille. Le passage de la topologie (mapping) à la réalisation sera qualifié de placement-routage" physique (le terme "placement" veut dire de placer les sommets de ce graphe donné sur une grille 2D et le terme "routage" veut dire de faire communiquer ces sommets entre eux)
Graphs with a minimum diameter have applications in the design of building-block switching systems, communication networks, and distributed computer systems. Several methods of constructing directed graphs with a small diameter are proposed. First, the dissertation presents as background several (delta, D) graphs including the Hypercube and de Bruijn. It shows the major disadvantages when implementing these topologies in practice for large scale. To achieve our goal, we propose a regular graph called Small World Heuristic (SWH) suitable for large parallel computers. This graph has a maximum degree ! and a small diameter D, while maintaining an acceptable level of connectivity. We show that this heuristic can connect on short distance thousands of nodes as little as 4 links per node. ̕Finally, we present a new integrated placement and routing algorithm to implement this heuristic on 2D VLSI

Style APA, Harvard, Vancouver, ISO itp.

42

Hamon, Ronan. "Analyse de réseaux temporels par des méthodes de traitement du signal : application au système de vélos en libre-service à Lyon". Thesis, Lyon, École normale supérieure, 2015. http://www.theses.fr/2015ENSL1017/document.

Pełny tekst źródła

Streszczenie:

Les systèmes de vélos en libre-service sont devenus des éléments indispensables dans les offres de transport urbain des grandes villes mondiales. À partir des données que ces systèmes génèrent, il est possible d'avoir une caractérisation fine de l'utilisation du vélo en milieu urbain, tant sur des problématiques traitant du domaine des transports que des aspects socio-économiques. Comme pour de nombreux domaines profitant de la récente abondance en données permises par les technologies actuelles de communication et de stockage de l'information, les enjeux actuels résident dans le développement de méthodes d'analyse de données efficaces et adaptées aux systèmes étudiés. Cette thèse se propose de répondre à cette problématique, à la fois par des développements méthodologiques et par une application à des données réelles issues du système de vélos en libre-service Vélo'v à Lyon.Le système Vélo'v peut se représenter sous la forme d'un réseau, décrivant un ensemble de relations entre les différentes stations. Cette représentation, valable également pour de nombreux systèmes, permet l'utilisation d'outils pour décrire la structure du réseau basés sur la théorie des graphes. Néanmoins, la prise en compte d'une dynamique temporelle dans l'évolution des systèmes nécessite d'étendre l'analyse à des réseaux temporels, dont la structure évolue au cours du temps. Le parallèle avec le domaine du traitement du signal, dont le but est l'analyse de signaux temporels, amène à considérer des connexions entre la description de l'évolution d'un réseau temporel et celle d'un signal. Ces travaux proposent de considérer une dualité entre les réseaux temporels et les signaux, de sorte que l'analyse dans le domaine des signaux, à l'aide des outils du traitement du signal, permet de caractériser le réseau temporel correspondant.Cette méthodologie, à la frontière entre le traitement du signal et l'analyse des réseaux, est tout d'abord justifiée par l'étude du système Vélo'v, en comparant différentes approches d'analyse de données et les apports de la représentation sous la forme de réseau temporel. Une méthode d'étiquetage des noeuds d'un graphe est ensuite discutée, permettant d'ouvrir la voie vers une dualité entre réseaux et signaux. Cette dualité est étendue aux réseaux temporels, pour lesquels une méthode d'extraction automatique des structures pertinentes au cours du temps est proposée, à travers la décomposition des signaux correspondants
Bike-sharing systems have become essential elements in urban transportation systems of many world's big cities. Thanks to the data generated by these systems, it is possible to obtain a precise characterization of urban cycling, both in terms of transportation and socio-economic aspects. Taking advantage of the recent abundance of data allowed by the current technology, the challenges lie in the development of efficient data analysis method, adapted to these systems. This PhD thesis proposes some answers to this issue, first by methodological developments and second by studying real-world data obtained from the bike-sharing system in Lyon, called Vélo'v.The Vélo'v system can be represented as a network, describing a set of relations between the stations spread over the city. This representation, used for many systems, enables the use of tools from network theory to measure the network structure and understand the underlying mechanisms. Nevertheless, taking into account the dynamic evolution of the structure requires an extension of the classical tools to the temporal case. Parallels between this problem and the field of signal processing can be done, and opens the way to the consideration of connections between the description of the dynamics of temporal networks and those of signals. This work introduces a duality between temporal networks and signals, such that the analysis of the signals using the classical tools of signal processing helps to the characterization of the structure of the corresponding network.This methodology, at the juncture between signal processing and network analysis, is first justified by the study of the Vélo'v network, by comparing different data analysis method and the representation of the system as a temporal network. Then, a method to relabel the vertices of the graph according to the topology of the network is discussed, opening up a duality between networks and signals. This duality is then extended to temporal networks: The analysis of the spectral properties of the signals are studied through a fully automated extraction method, enabling the decomposition of relevant network structure over time

Style APA, Harvard, Vancouver, ISO itp.

43

Hain, Horst-Udo. "Phonetische Transkription für ein multilinguales Sprachsynthesesystem". Doctoral thesis, Saechsische Landesbibliothek- Staats- und Universitaetsbibliothek Dresden, 2012. http://nbn-resolving.de/urn:nbn:de:bsz:14-qucosa-81777.

Pełny tekst źródła

Streszczenie:

Die vorliegende Arbeit beschäftigt sich mit einem datengetriebenen Verfahren zur Graphem-Phonem-Konvertierung für ein Sprachsynthesesystem. Die Aufgabe besteht darin, die Aussprache für beliebige Wörter zu bestimmen, auch für solche Wörter, die nicht im Lexikon des Systems enthalten sind. Die Architektur an sich ist sprachenunabhängig, von der Sprache abhängig sind lediglich die Wissensquellen, die zur Laufzeit des Systems geladen werden. Die Erstellung von Wissensquellen für weitere Sprachen soll weitgehend automatisch und ohne Einsatz von Expertenwissen möglich sein. Expertenwissen kann verwendet werden, um die Ergebnisse zu verbessern, darf aber keine Voraussetzung sein. Für die Bestimmung der Transkription werden zwei neuronale Netze verwendet. Das erste Netz generiert aus der Buchstabenfolge des Wortes die zu realisierenden Laute einschließlich der Silbengrenzen, und das zweite bestimmt im Anschluß daran die Position der Wortbetonung. Diese Trennung hat den Vorteil, daß man für die Bestimmung des Wortakzentes das Wissen über die gesamte Lautfolge einbeziehen kann. Andere Verfahren, die die Transkription in einem Schritt bestimmen, haben das Problem, bereits zu Beginn des Wortes über den Akzent entscheiden zu müssen, obwohl die Aussprache des Wortes noch gar nicht feststeht. Zudem bietet die Trennung die Möglichkeit, zwei speziell auf die Anforderung zugeschnittene Netze zu trainieren. Die Besonderheit der hier verwendeten neuronalen Netze ist die Einführung einer Skalierungsschicht zwischen der eigentlichen Eingabe und der versteckten Schicht. Eingabe und Skalierungsschicht werden über eine Diagonalmatrix verbunden, wobei auf die Gewichte dieser Verbindung ein Weight Decay (Gewichtezerfall) angewendet wird. Damit erreicht man eine Bewertung der Eingabeinformation während des Trainings. Eingabeknoten mit einem großen Informationsgehalt werden verstärkt, während weniger interessante Knoten abgeschwächt werden. Das kann sogar soweit gehen, daß einzelne Knoten vollständig abgetrennt werden. Der Zweck dieser Verbindung ist, den Einfluß des Rauschens in den Trainingsdaten zu reduzieren. Durch das Ausblenden der unwichtigen Eingabewerte ist das Netz besser in der Lage, sich auf die wichtigen Daten zu konzentrieren. Das beschleunigt das Training und verbessert die erzielten Ergebnisse. In Verbindung mit einem schrittweisen Ausdünnen der Gewichte (Pruning) werden zudem störende oder unwichtige Verbindungen innerhalb der Netzwerkarchitektur gelöscht. Damit wird die Generalisierungsfähigkeit noch einmal erhöht. Die Aufbereitung der Lexika zur Generierung der Trainingsmuster für die neuronalen Netze wird ebenfalls automatisch durchgeführt. Dafür wird mit Hilfe der dynamischen Zeitanpassung (DTW) der optimale Pfad in einer Ebene gesucht, die auf der einen Koordinate durch die Buchstaben des Wortes und auf der anderen Koordinate durch die Lautfolge aufgespannt wird. Somit erhält man eine Zuordnung der Laute zu den Buchstaben. Aus diesen Zuordnungen werden die Muster für das Training der Netze generiert. Um die Transkriptionsergebnisse weiter zu verbessern, wurde ein hybrides Verfahren unter Verwendung der Lexika und der Netze entwickelt. Unbekannte Wörter werden zuerst in Bestandteile aus dem Lexikon zerlegt und die Lautfolgen dieser Teilwörter zur Gesamttranskription zusammengesetzt. Dabei werden Lücken zwischen den Teilwörtern durch die neuronalen Netze aufgefüllt. Dies ist allerdings nicht ohne weiteres möglich, da es zu Fehlern an den Schnittstellen zwischen den Teiltranskriptionen kommen kann. Dieses Problem wird mit Hilfe des Lexikons gelöst, das für die Generierung der Trainingsmuster aufbereitet wurde. Hier ist eine eindeutige Zuordnung der Laute zu den sie generierenden Buchstaben enthalten. Somit können die Laute an den Schnittstellen neu bewertet und Transkriptionsfehler vermieden werden. Die Verlagsausgabe dieser Dissertation erschien 2005 im w.e.b.-Universitätsverlag Dresden (ISBN 3-937672-76-1)
The topic of this thesis is a system which is able to perform a grapheme-to-phoneme conversion for several languages without changes in its architecture. This is achieved by separation of the language dependent knowledge bases from the run-time system. Main focus is an automated adaptation to new languages by generation of new knowledge bases without manual effort with a minimal requirement for additional information. The only source is a lexicon containing all the words together with their appropriate phonetic transcription. Additional knowledge can be used to improve or accelerate the adaptation process, but it must not be a prerequisite. Another requirement is a fully automatic process without manual interference or post-editing. This allows for the adaptation to a new language without even having a command of that language. The only precondition is the pronunciation dictionary which should be enough for the data-driven approach to learn a new language. The automatic adaptation process is divided into two parts. In the first step the lexicon is pre-processed to determine which grapheme sequence belongs to which phoneme. This is the basis for the generation of the training patterns for the data-driven learning algorithm. In the second part mapping rules are derived automatically which are finally used to create the phonetic transcription of any word, even if it not contained in the dictionary. Task is to have a generalisation process that can handle all words in a text that has to be read out by a text-to-speech system

Style APA, Harvard, Vancouver, ISO itp.

44

Mimeur, Christophe. "Les traces de la vitesse entre réseau et territoire : approche géohistorique de la croissance du réseau ferroviaire français". Thesis, Dijon, 2016. http://www.theses.fr/2016DIJOL028/document.

Pełny tekst źródła

Streszczenie:

Les interactions entre transport et territoire sont l’objet d’une littérature scientifique permanente, questionnant les impacts économiques et démographiques d’une nouvelle infrastructure, souvent évoqués à l’échelle d’un projet. L’objectif de la thèse est de réinvestir les composantes de l’interaction par les larges échelles spatiales et temporelles, en posant l’hypothèse que la profondeur temporelle et l’échelle du territoire national sont porteuses de nouvelles explications. Ce travail s’appuie sur la collecte, l’exploitation et l’analyse de la large base de données FRANcE (French RAilway NEtwork), qui recense chaque section du réseau ferroviaire français depuis le début du XIXème siècle et les recensements démographiques. Cette base renferme également les traces de la vitesse, qui constituent une information inédite sur l’ensemble du réseau et qui permet de faire de l’accessibilité une variable décisive dans les explications. Plutôt que de se concentrer sur l’acquisition de nouvelles données au prix d’une lourde collecte, nous misons sur la construction d’un appareil méthodologique pour étudier les deux sens de l’interaction entre réseau et territoire, qui requiert toutefois une adaptation des dispositifs de structuration des données et d’analyse. La démarche de la thèse consiste en une modélisation croissante du phénomène, de la compréhension et la formalisation des objets jusqu’à la formalisation des données et des analyses, ce qui nécessite le recours à d’autres disciplines. Ce travail utilise le formalisme des graphes pour investiguer les deux sens de la relation. Il aide à étudier l’effet du réseau à partir d’une diversification de la donnée et de sa modélisation pour rendre compte de portées spatiales et temporelles. Il aide à étudier l’impact d’une structure préexistante dans la morphogénèse du réseau ferroviaire français à partir d’un modèle d’évolution endogène, entre diffusion du rail et hiérarchisation des infrastructures. Ce travail vise à mieux comprendre les liens qui unissent réseau et territoire, dont les outils méthodologiques peuvent être appliquées à d’autres réseaux, d’autres temporalités, jusqu’à des problématiques actuelles
The interaction between space and network are frequently questioned in the academic literature, by asking the economical and demographical impacts of a new infrastructure, often studied at the scale of a project. This work aims to investigate the components of the interaction in both large spatial and temporal scales. The hypothesis is that the temporal depth and the national scale could bring new explanations. This work is based on the collect, the exploitation and the analysis of the large spatio-temporal database FRANcE (French Railway Network). It identifies all sections of the network since the 19th century and the population census. This database also contains the traces of the speed, which are novel information for network, and allows the accessibility to become a decisive variable in the explanations. Rather than acquisition new data with an intensive phase of collect, we aim to build a methodological chain to study the two senses of interaction between space and network. It requires the adaptation of data structuration and analysis. The approach of this thesis consists on the growing modelling of the phenomenon, from the comprehension to formalization of data to the analysis, which requires the use of other disciplines. This work uses the graph theory to investigate the two senses of the relationship. It permits to study the network effect in the long run by diversifying the data to identify spatial and temporal ranges. It permits to study the impact of a pre-existing structure in the morphogenesis of the network, by using a dynamic model of network evolution, between diffusion and hierarchical organization. This work aims to understand the link between space and network, where the methodological tools can be adapted to other networks, other times and actual questioning

Style APA, Harvard, Vancouver, ISO itp.

45

Hollocou, Alexandre. "Nouvelles approches pour le partitionnement de grands graphes". Thesis, Paris Sciences et Lettres (ComUE), 2018. http://www.theses.fr/2018PSLEE063.

Pełny tekst źródła

Streszczenie:

Les graphes sont omniprésents dans de nombreux domaines de recherche, allant de la biologie à la sociologie. Un graphe est une structure mathématique très simple constituée d’un ensemble d’éléments, appelés nœuds, reliés entre eux par des liens, appelés arêtes. Malgré cette simplicité, les graphes sont capables de représenter des systèmes extrêmement complexes, comme les interactions entre protéines ou les collaborations scientifiques. Le partitionnement ou clustering de graphe est un problème central en analyse de graphe dont l’objectif est d’identifier des groupes de nœuds densément interconnectés et peu connectés avec le reste du graphe. Ces groupes de nœuds, appelés clusters, sont fondamentaux pour une compréhension fine de la structure des graphes. Il n’existe pas de définition universelle de ce qu’est un bon cluster, et différentes approches peuvent s’avérer mieux adaptées dans différentes situations. Alors que les méthodes classiques s’attachent à trouver des partitions des nœuds de graphe, c’est-à-dire à colorer ces nœuds de manière à ce qu’un nœud donné n’ait qu’une et une seule couleur, des approches plus élaborées se révèlent nécessaires pour modéliser la structure complexe des graphes que l’on rencontre en situation réelle. En particulier, dans de nombreux cas, il est nécessaire de considérer qu’un nœud donné peut appartenir à plus d’un cluster. Par ailleurs, de nombreux systèmes que l’on rencontre en pratique présentent une structure multi-échelle pour laquelle il est nécessaire de partir à la recherche de hiérarchies de clusters plutôt que d’effectuer un partitionnement à plat. De plus, les graphes que l’on rencontre en pratique évoluent souvent avec le temps et sont trop massifs pour être traités en un seul lot. Pour ces raisons, il est souvent nécessaire d’adopter des approches dites de streaming qui traitent les arêtes au fil de l’eau. Enfin, dans de nombreuses applications, traiter des graphes entiers n’est pas nécessaire ou est trop coûteux, et il est plus approprié de retrouver des clusters locaux dans un voisinage de nœuds d’intérêt plutôt que de colorer tous les nœuds. Dans ce travail, nous étudions des approches alternatives de partitionnement de graphe et mettons au point de nouveaux algorithmes afin de résoudre les différents problèmes évoqués ci-dessus
Graphs are ubiquitous in many fields of research ranging from sociology to biology. A graph is a very simple mathematical structure that consists of a set of elements, called nodes, connected to each other by edges. It is yet able to represent complex systems such as protein-protein interaction or scientific collaborations. Graph clustering is a central problem in the analysis of graphs whose objective is to identify dense groups of nodes that are sparsely connected to the rest of the graph. These groups of nodes, called clusters, are fundamental to an in-depth understanding of graph structures. There is no universal definition of what a good cluster is, and different approaches might be best suited for different applications. Whereas most of classic methods focus on finding node partitions, i.e. on coloring graph nodes so that each node has one and only one color, more elaborate approaches are often necessary to model the complex structure of real-life graphs and to address sophisticated applications. In particular, in many cases, we must consider that a given node can belong to more than one cluster. Besides, many real-world systems exhibit multi-scale structures and one much seek for hierarchies of clusters rather than flat clusterings. Furthermore, graphs often evolve over time and are too massive to be handled in one batch so that one must be able to process stream of edges. Finally, in many applications, processing entire graphs is irrelevant or expensive, and it can be more appropriate to recover local clusters in the neighborhood of nodes of interest rather than color all graph nodes. In this work, we study alternative approaches and design novel algorithms to tackle these different problems. The novel methods that we propose to address these different problems are mostly inspired by variants of modularity, a classic measure that accesses the quality of a node partition, and by random walks, stochastic processes whose properties are closely related to the graph structure. We provide analyses that give theoretical guarantees for the different proposed techniques, and endeavour to evaluate these algorithms on real-world datasets and use cases

Style APA, Harvard, Vancouver, ISO itp.

46

Leichtnam, Laetitia. "Detecting and visualizing anomalies in heterogeneous network events : Modeling events as graph structures and detecting communities and novelties with machine learning". Thesis, CentraleSupélec, 2020. http://www.theses.fr/2020CSUP0011.

Pełny tekst źródła

Streszczenie:

L'objectif général de cette thèse est d'évaluer l'intérêt des graphes dans le domaine de l'analyse des données de sécurité.Nous proposons une approche de bout en bout composé d'un modèle unifié de données réseau sous forme de graphes, d'un système de découverte de communauté, d'un système de détection d'anomalies non supervisé et d'une visualisation des données sous forme de graphes. Le modèle unifié est obtenue en utilisant des graphes de connaissance pour représenter des journaux d'évènements hétérogènes ainsi que du trafic réseau. La détection de communautés permet de sélectionner des sous-graphes représentant des événements fortement liés à une alerte ou à un IoC et qui sont donc pertinents pour l'analyse forensique. Notre système de détection d'intrusion basé sur les anomalies repose sur la détection de nouveauté par un autoencodeur et donne de très bons résultats sur les jeux de données CICIDS 2017 et 2018. Enfin, la visualisation immersive des données de sécurité permet de mettre en évidence les relations entre les éléments de sécurité et les événements malveillants ou les IoCs. Cela donne à l'analyste de sécurité un bon point de départ pour explorer les données et reconstruire des scénarii d'attaques globales
The general objective of this thesis is to evaluate the interest of graph structures in the field of security data analysis.We propose an end-to-end approach consisting in a unified view of the network data in the form of graphs, a community discovery system, an unsupervised anomaly detection system, and a visualization of the data in the form of graphs. The unified view is obtained using knowledge graphs to represent heterogeneous log files and network traffics. Community detection allows us to select sub-graphs representing events that are strongly related to an alert or an IoC and that are thus relevant for forensic analysis. Our anomaly-based intrusion detection system relies on novelty detection by an autoencoder and exhibits very good results on CICIDS 2017 and 2018 datasets. Finally, an immersive visualization of security data allows highlighting the relations between security elements and malicious events or IOCs. This gives the security analyst a good starting point to explore the data and reconstruct global attack scenarii

Style APA, Harvard, Vancouver, ISO itp.

47

Chelly, Magda Lilia. "Propagation d'une position dans les réseaux connectés". Phd thesis, Institut National des Télécommunications, 2011. http://tel.archives-ouvertes.fr/tel-00843587.

Pełny tekst źródła

Streszczenie:

Les systèmes de positionnement ont connu un progrès indéniable. Actuellement, la précision atteint quelques centimètres sous certaines conditions : espace ouvert, ciel dégagé, technique très spécifique, etc. Néanmoins, le problème du positionnement dans un environnement intérieur demeure persistant : les trajets multiples qui compliquent les modèles de propagation, l'atténuation, etc. Différents systèmes ont vu le jour, utilisant des technologies telles que l'UWB, le WiFi ou l'Infrarouge. Ces systèmes apportent des résultats de positionnement intéressants, atteignant l'ordre du mètre. Cette précision reste liée à certaines contraintes : une infrastructure, une technologie utilisée, une calibration, une technique de calcul, etc. Afin de réduire toutes ces contraintes, nous proposons une nouvelle approche de positionnement. Notre approche utilise tous les équipements réseaux présents dans un environnement. Elle se base sur deux étapes fondamentales : l'étude de visibilité et l'élaboration de liens géographiques. L'étude de visibilité permet d'obtenir les équipements visibles par un équipement. Nous avons exposé plusieurs modèles de visibilité et nous avons effectué une comparaison des résultats. L'élaboration de liens géographiques permet de construire un graphe géographique tridimensionnel reliant tous les équipements de l'environnement. Ce graphe nous permet de visualiser la répartition des équipements et d'estimer les positions géographiques de chaque équipement. Pour la mise en œuvre de notre approche, nous avons développé un simulateur sous Matlab. Le simulateur élaboré évalue d'abord le nombre d'équipements visibles. Il estime les distances séparant cet équipement de chaque équipement visible. Enfin, il construit un graphe géographique et calcul les positions géographiques. Des résultats de simulations sont présentés pour valider notre approche qui permet d'aboutir à un système capable d'opérer, sans aucune infrastructure additionnelle, un positionnement dans un environnement intérieur et extérieur

Style APA, Harvard, Vancouver, ISO itp.

48

Martineau, Maxime. "Deep learning onto graph space : application to image-based insect recognition". Thesis, Tours, 2019. http://www.theses.fr/2019TOUR4024.

Pełny tekst źródła

Streszczenie:

Le but de cette thèse est d'étudier la reconnaissance d'insectes comme un problème de reconnaissance des formes basé images. Bien que ce problème ait été étudié en profondeur au long des trois dernières décennies, un aspect reste selon nous toujours à expérimenter à ce jour : les approches profondes (deep learning). À cet effet, la première contribution de cette thèse consiste à déterminer la faisabilité de l'application des réseaux de neurones convolutifs profonds (CNN) au problème de reconnaissance d'images d'insectes. Les limitations majeures ont les suivantes: les images sont très rares et les cardinalités de classes sont hautement déséquilibrées. Pour atténuer ces limitations, le transfer learning et la pondération de la fonction de coûts ont été employés. Des méthodes basées graphes sont également proposées et testées. La première consiste en la conception d'un classificateur de graphes de type perceptron. Le second travail basé sur les graphes de cette thèse est la définition d'un opérateur de convolution pour construire un modèle de réseaux de neurones convolutifs s'appliquant sur les graphes (GCNN.) Le dernier chapitre de la thèse s'applique à utiliser les méthodes mentionnées précédemment à des problèmes de reconnaissance d'images d'insectes. Deux bases d'images sont ici proposées. Là première est constituée d'images prises en laboratoire sur arrière-plan constant. La seconde base est issue de la base ImageNet. Cette base est composée d'images prises en contexte naturel. Les CNN entrainés avec transfer learning sont les plus performants sur ces bases d'images
The goal of this thesis is to investigate insect recognition as an image-based pattern recognition problem. Although this problem has been extensively studied along the previous three decades, an element is to the best of our knowledge still to be experimented as of 2017: deep approaches. Therefore, a contribution is about determining to what extent deep convolutional neural networks (CNNs) can be applied to image-based insect recognition. Graph-based representations and methods have also been tested. Two attempts are presented: The former consists in designing a graph-perceptron classifier and the latter graph-based work in this thesis is on defining convolution on graphs to build graph convolutional neural networks. The last chapter of the thesis deals with applying most of the aforementioned methods to insect image recognition problems. Two datasets are proposed. The first one consists of lab-based images with constant background. The second one is generated by taking a ImageNet subset. This set is composed of field-based images. CNNs with transfer learning are the most successful method applied on these datasets

Style APA, Harvard, Vancouver, ISO itp.

49

Solouki, Bonab Vahab. "Polyurethane (PU) Nanocomposites; Interplay of Composition, Morphology, and Properties". Case Western Reserve University School of Graduate Studies / OhioLINK, 2019. http://rave.ohiolink.edu/etdc/view?acc_num=case1542634359353501.

Pełny tekst źródła

Style APA, Harvard, Vancouver, ISO itp.

50

Issartel, Yann. "Inférence sur des graphes aléatoires". Thesis, université Paris-Saclay, 2020. http://www.theses.fr/2020UPASM019.

Pełny tekst źródła

Streszczenie:

Cette thèse s'inscrit dans les domaines de la statistique non-paramétrique et de la théorie statistique de l'apprentissage non-supervisé. Son objet est la compréhension et la mise en œuvre de méthodes d'estimation et de décision pour des modèles de graphes aléatoires à espace latent. Ces outils probabilistes rencontrent un succès grandissant pour la modélisation de grands réseaux dans des domaines aussi différents que la biologie, le marketing ou les sciences sociales. Dans un premier temps, nous définissons un indice identifiable de la dimension de l'espace latent puis un estimateur consistant de cet indice. Plus généralement, ces quantités identifiables et interprétables permettent de palier l'absence d'identifiabilité de l'espace latent lui-même. Dans la suite, nous introduisons le problème de `pair-matching'. En partant d'un graphe non-observé, une stratégie choisit de façon séquentielle des paires de nœuds et observe la présence/absence d'arêtes. Son objectif est de découvrir le plus grand nombre possible d'arêtes avec un budget fixé. Pour ce problème de type bandit, nous étudions les regrets optimaux dans un modèle à blocs stochastiques puis dans un graphe aléatoire géométrique. Enfin, nous estimons les positions des nœuds dans l'espace latent, dans le cas particulier où l'espace est un cercle dans le plan euclidien. Pour chacun des trois problèmes, nous obtenons des procédures optimales au sens minimax, ainsi que des procédures efficaces satisfaisant certaines garanties théoriques. Ces algorithmes sont analysés d'un point de vue non-asymptotique en s'appuyant, entre autres, sur des inégalités de concentration
This thesis lies at the intersection of the theories of non-parametric statistics and statistical learning. Its goal is to provide an understanding of statistical problems in latent space random graphs. Latent space models have emerged as useful probabilistic tools for modeling large networks in various fields such as biology, marketing or social sciences. We first define an identifiable index of the dimension of the latent space and then a consistent estimator of this index. More generally, such identifiable and interpretable quantities alleviate the absence of identifiability of the latent space itself. We then introduce the pair-matching problem. From a non-observed graph, a strategy sequentially queries pairs of nodes and observes the presence/absence of edges. Its goal is to discover as many edges as possible with a fixed budget of queries. For this bandit type problem, we study optimal regrets in stochastic block models and random geometric graphs. Finally, we are interested in estimating the positions of the nodes in the latent space, in the particular situation where the space is a circle in the Euclidean plane. For each of the three problems, we obtain procedures that achieve the statistical optimal performance, as well as efficient procedures with theoretical guarantees. These algorithms are analysed from a non-asymptotic viewpoint, relying in particular on concentration inequalities

Style APA, Harvard, Vancouver, ISO itp.

Rozprawy doktorskie na temat „Graphyne networks”

Utwórz poprawne odniesienie w stylach APA, MLA, Chicago, Harvard i wielu innych