Contents
Academic literature on the topic 'Sélection topologique'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the lists of relevant articles, books, theses, conference reports, and other scholarly sources on the topic 'Sélection topologique.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Journal articles on the topic "Sélection topologique"
BERTRAND, Denis. "La générativité est-elle soluble dans le sensible ? Réflexions topologiques et énonciatives « au cœur » du parcours génératif." 130, no. 130 (January 23, 2024). http://dx.doi.org/10.25965/as.8295.
Full textDissertations / Theses on the topic "Sélection topologique"
Simoncini, David. "Sélection topologique dans les algorithmes évolutionnaires cellulaires : étude du compromis exploration exploitation." Nice, 2009. http://www.theses.fr/2009NICE4079.
Full textEvolutionary algorithms are stochastic optimization methods manipulating a population of solutions. Their behaviour is inspired by Darwin's theory of evolution. The combined application of stochastic operators and selection mechanisms allow renewing the population by exploring the search space and exploiting the already found solutions. The convergence speed of an evolutionary algorithm relies on its ability to generate efficient solutions by leading the search toward promising regions of the search space, and the ability of solutions to survive according to their fitness defined by the selective pressure. The latter allows dealing with the exploration / exploitation trade-off and prevents the algorithm from converging prematurely toward a local optimum. Evolutionary cellular algorithms introduce a notion of geographical neighborhood by embedding the solution on a grid. This adds a topological level between the phenotypical and genotypical ones. In this context, we define new selection methods that allow controlling the topology and obtain complex dynamics thanks to a single continuous and bounded parameter. Instead of restricting solutions to evolve on a uniform grid, we propose to enhance the topology with notions of anisotropy and locality. We study the influence of the topological selection on the preservation of genotypic diversity. Experiences made on two classes of NP-complete problems show that taking into account the topological level leads to a fine equilibrium between exploration and exploitation. In order to study the search dynamic and especially to analyze the efficiency of the observed trade-offs, we define a model based on the notion of punctuated equilibria. Finally, we propose adaptive algorithms in the intent of dynamically controlling the selective pressure and thus dealing with the relation between exploration and exploitation phases without any knowledge on the studied problems
Doan, Nath-Quang. "Modèles hiérarchiques et topologiques pour le clustering et la visualisation des données." Paris 13, 2013. http://scbd-sto.univ-paris13.fr/secure/edgalilee_th_2013_doan.pdf.
Full textThis thesis focuses on clustering approaches inspired from topological models and an autonomous hierarchical clustering method. The clustering problem becomes more complicated and difficult due to the growth in quality and quantify of structured data such as graphs, trees or sequences. In this thesis, we are particularly interested in self-organizing maps which have been generally used for learning topological preservation, clustering, vector quantization and graph visualization. Our studyconcerns also a hierarchical clustering method AntTree which models the ability of real ants to build structure by connect themselves. By combining the topological map with the self-assembly rules inspired from AntTree, the goal is to represent data in a hierarchical and topological structure providing more insight data information. The advantage is to visualize the clustering results as multiple hierarchical trees and a topological network. In this report, we present three new models that are able to address clustering, visualization and feature selection problems. In the first model, our study shows the interest in the use of hierarchical and topological structure through several applications on numerical datasets, as well as structured datasets e. G. Graphs and biological dataset. The second model consists of a flexible and growing structure which does not impose the strict network-topology preservation rules. Using statistical characteristics provided by hierarchical trees, it accelerates significantly the learning process. The third model addresses particularly the issue of unsupervised feature selection. The idea is to use hierarchical structure provided by AntTree to discover automatically local data structure and local neighbors. By using the tree topology, we propose a new score for feature selection by constraining the Laplacian score. Finally, this thesis offers several perspectives for future work
Aazi, Fatima Zahra. "Contribution à la sélection de variables par les machines à vecteurs support pour la discrimination multi-classes." Thesis, Lyon, 2016. http://www.theses.fr/2016LYSE2180.
Full textThe technological progress has allowed the storage of large amounts of data in terms of size (number of observations) and dimensions (number of variables). These data require new methods, especially for predictive modeling (data science), of statistical processing adapted to their characteristics. In this thesis, we are particularly interested in the data with large numberof variables compared to the number of observations.For these data, reducing the number of initial variables, hence dimensions, by selecting an optimal subset is necessary, even imperative. It reduces the complexity, helps to understand the data structure, improves the interpretation of the results and especially enhances the performance of the forecasting model by eliminating redundant and / or noise variables.More precisely, we are interested in the selection of variables in the context of supervised learning, specifically of multiclass discrimination. The objective is to propose some new methods of variable selection for multiclass discriminant models called Multiclass Support Vector Machines (MSVM).Two approaches are proposed in this work. The first one, presented in a classical context, consist in selecting the optimal subset of variables using the radius margin upper bound of the generalization error of MSVM. The second one, proposed in a topological context, uses the concepts of neighborhood graphs and the degree of topological equivalence in discriminationto identify the relevant variables and to select the optimal subset for an MSVM model.The evaluation of these two approaches on simulated and real data shows that they can select from a large number of initial variables, a reduced number providing equal or better performance than those obtained by competing methods
Yeresko, Mykhailo. "Search for antineutrino disappearance with the SoLi∂ detector : novel reconstruction, calibration and selection." Electronic Thesis or Diss., Université Clermont Auvergne (2021-...), 2022. http://www.theses.fr/2022UCFAC127.
Full textThis thesis presents the search for active-to-sterile antineutrino oscillations with the SoLi∂ detector, which has an innovative concept based on a sandwich of composite polyvinyl-toluene and 6LiF:ZnS(Ag) scintillators. Three main contributions to this search are reported. The first one describes the new reconstruction method based on the ML-EM algorithm. It is aimed at transformation of the raw detector response to the list of positions (with dedicated energies) where actual physics interaction in the detector took place. The second one describes the energy calibration of the detector. It includes both relative calibration with the cosmic horizontal muons and exploration of several options for the absolute energy scale derivation. Finally, the thesis presents a novel method for selecting antineutrino candidates based on the analysis of the electromagnetic part of the inverse beta decay signal and categorising them according to the geometry of the events in the detector (topologies). The analysis was developed blindly and the method was validated with a small fraction of the data sample
Dhifli, Wajdi. "Fouille de Sous-graphes Basée sur la Topologie et la Connaissance du Domaine: Application sur les Structures 3D de Protéines." Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2013. http://tel.archives-ouvertes.fr/tel-00922209.
Full textLejosne, Yohan. "Quelques aspects des réseaux multi-cellules multi-utilisateurs MIMO : délai, conception d'émetteur-récepteur, sélection d'utilisateurs et topologie." Thesis, Paris, ENST, 2014. http://www.theses.fr/2014ENST0089/document.
Full textIn order to meet ever-growing needs for capacity in wireless networks, transmission techniques and the system models used to study their performances have rapidly evolved. From single-user single-antenna point-to-point communications to modern multi-cell multi-antenna networks there have been large advances in technology. Along the way, several assumptions are made in order to have either more realistic models, but also to allow simpler analysis. We analyze three aspects of actual networks and try to benefit from them when possible or conversely, to mitigate their negative impact. We focus on the delay in the CSI acquisition. Precisely, when taken into account, this delay greatly impairs the channel multiplexing gain if nothing is done to use the dead time during which the transmitters are not transmitting and do not yet have the CSI. We review and propose different schemes to efficiently use this dead time to improve the multiplexing gain in both the BC and the interference channel (IC). We evaluate the more relevant net multiplexing gain, taking intoaccount the training and feedback overhead. Results are surprising because potential schemes to fight delay reveal to be burdened byimpractical overheads in the BC. In the IC, an optimal scheme is proposed. It allows avoiding any loss of multiplexing gain even forsignificant delay in the CSI acquisition. Concerning the number of users, we propose a new criterion for the greedy user selection in a BC to benefit of the multi-user diversity, and two interference alignment schemes for the IC to benefit of having multiple users in each cell. Finally, partially connected cellular networks are considered and schemes to benefit from said partial connectivity to increase the multiplexing gain are proposed
Dhifli, Wajdi. "Topological and domain Knowledge-based subgraph mining : application on protein 3D-structures." Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2013. http://tel.archives-ouvertes.fr/tel-00946989.
Full textLafite, Pierre. "ETUDE du CYTOCHROME P450 2J2 HUMAIN :Recherche de substrats et d'inhibiteurs sélectifs ;Détermination de la topologie de son site actif." Phd thesis, Université René Descartes - Paris V, 2007. http://tel.archives-ouvertes.fr/tel-00192090.
Full textLedoux, Audrey. "Vers des traitements morphologiques couleur et spectraux valides au sens perceptuel et physique : Méthodes et critères de sélection." Phd thesis, Université de Poitiers, 2013. http://tel.archives-ouvertes.fr/tel-00965767.
Full textDilmi, Mohamed Djallel. "Méthodes de classification des séries temporelles : application à un réseau de pluviomètres." Electronic Thesis or Diss., Sorbonne université, 2019. https://accesdistant.sorbonne-universite.fr/login?url=https://theses-intra.sorbonne-universite.fr/2019SORUS087.pdf.
Full textThe impact of climat change on the temporal evolution of precipitation as well as the impact of the Parisian heat island on the spatial distribution of précipitation motivate studying the varaibility of the water cycle on a small scale on île-de-france. one way to analyse this varaibility using the data from a rain gauge network is to perform a clustring on time series measured by this network. In this thesis, we have explored two approaches for time series clustring : for the first approach based on the description of series by characteristics, an algorithm for selecting characteristics based on genetic algorithms and topological maps has been proposed. for the second approach based on shape comparaison, a measure of dissimilarity (iterative downscaling time warping) was developed to compare two rainfall time series. Then the limits of the two approaches were discuddes followed by a proposition of a mixed approach that combine the advantages of each approach. The approach was first applied to the evaluation of spatial variability of precipitation on île-de-france. For the evaluation of the temporal variability of the precpitation, a clustring on the precipitation events observed by a station was carried out then extended on the whole rain gauge network. The application on the historical series of Paris-Montsouris (1873-2015) makes it possible to automatically discriminate "remarkable" years from a meteorological point of view