Rozprawy doktorskie na temat „Réseaux de neurones pour graphes”
Utwórz poprawne odniesienie w stylach APA, MLA, Chicago, Harvard i wielu innych
Sprawdź 50 najlepszych rozpraw doktorskich naukowych na temat „Réseaux de neurones pour graphes”.
Przycisk „Dodaj do bibliografii” jest dostępny obok każdej pracy w bibliografii. Użyj go – a my automatycznie utworzymy odniesienie bibliograficzne do wybranej pracy w stylu cytowania, którego potrzebujesz: APA, MLA, Harvard, Chicago, Vancouver itp.
Możesz również pobrać pełny tekst publikacji naukowej w formacie „.pdf” i przeczytać adnotację do pracy online, jeśli odpowiednie parametry są dostępne w metadanych.
Przeglądaj rozprawy doktorskie z różnych dziedzin i twórz odpowiednie bibliografie.
Carboni, Lucrezia. "Graphes pour l’exploration des réseaux de neurones artificiels et de la connectivité cérébrale humaine". Electronic Thesis or Diss., Université Grenoble Alpes, 2023. http://www.theses.fr/2023GRALM060.
Pełny tekst źródłaThe main objective of this thesis is to explore brain and artificial neural network connectivity from agraph-based perspective. While structural and functional connectivity analysis has been extensivelystudied in the context of the human brain, there is a lack of a similar analysis framework in artificialsystems.To address this gap, this research focuses on two main axes.In the first axis, the main objective is to determine a healthy signature characterization of the humanbrain resting state functional connectivity. To achieve this objective, a novel framework is proposed,integrating traditional graph statistics and network reduction tools, to determine healthy connectivitypatterns. Hence, we build a graph pair-wise comparison and a classifier to identify pathological statesand rank associated perturbed brain regions. Additionally, the generalization and robustness of theproposed framework were investigated across multiple datasets and variations in data quality.The second research axis explores the benefits of brain-inspired connectivity exploration of artificialneural networks (ANNs) in the future perspective of more robust artificial systems development. Amajor robustness issue in ANN models is represented by catastrophic forgetting when the networkdramatically forgets previously learned tasks when adapting to new ones. Our work demonstrates thatgraph modeling offers a simple and elegant framework for investigating ANNs, comparing differentlearning strategies, and detecting deleterious behaviors such as catastrophic forgetting.Moreover, we explore the potential of leveraging graph-based insights to effectively mitigatecatastrophic forgetting, laying a foundation for future research and explorations in this area
Hérault, Laurent. "Réseaux de neurones récursifs pour l'optimisation combinatoire : application à la théorie des graphes et à la vision par ordinateur". Grenoble INPG, 1991. http://www.theses.fr/1991INPG0019.
Pełny tekst źródłaHafidi, Hakim. "Robust machine learning for Graphs/Networks". Electronic Thesis or Diss., Institut polytechnique de Paris, 2023. http://www.theses.fr/2023IPPAT004.
Pełny tekst źródłaThis thesis addresses advancements in graph representation learning, focusing on the challengesand opportunities presented by Graph Neural Networks (GNNs). It highlights the significanceof graphs in representing complex systems and the necessity of learning node embeddings that capture both node features and graph structure. The study identifies key issues in GNNs, such as their dependence on high-quality labeled data, inconsistent performanceacross various datasets, and susceptibility to adversarial attacks.To tackle these challenges, the thesis introduces several innovative approaches. Firstly, it employs contrastive learning for node representation, enabling self-supervised learning that reduces reliance on labeled data. Secondly, a Bayesian-based classifier isproposed for node classification, which considers the graph’s structure to enhance accuracy. Lastly, the thesis addresses the vulnerability of GNNs to adversarialattacks by assessing the robustness of the proposed classifier and introducing effective defense mechanisms.These contributions aim to improve both the performance and resilience of GNNs in graph representation learning
Boulnois, Philippe. "Contribution à l'étude de différentes architectures de réseaux de neurones artificiels réalisant une transcription graphèmes-phonèmes pour le français". Compiègne, 1994. http://www.theses.fr/1994COMPD675.
Pełny tekst źródłaHammadi, Youssef. "Réduction d'un modèle 0D instationnaire et non-linéaire de thermique habitacle pour l’optimisation énergétique des véhicules automobiles". Thesis, Université Paris sciences et lettres, 2020. http://www.theses.fr/2020UPSLM027.
Pełny tekst źródłaThe use of automotive air conditioning leads to a fuel overconsumption. To reduce this overconsumption, we can either work upstream on the technical definitions of the cabin and the HVAC system or optimize control strategies. In both cases, it is essential to build a cabin thermal model that well balances accuracy and complexity. This is the topic of this PhD thesis driven by Renault Group. First, a model reduction methodology is used to build a 0D model starting from a 3D finite element cabin thermal model. This 0D model is based on mass and energy balances on the different cabin walls and air zones. It consists of a nonlinear differential algebraic equations system which can be reinterpreted as a Bond Graph. In addition, the 0D model is based on a weak coupling between the thermal equations and the fluid mechanics ones resulting from CFD calculations (internal airflow and external aerodynamics). Secondly, we apply a machine learning method to the data generated by the 0D model in order to build a reduced 0D model. A design of experiment is considered at this stage. Due to the nonlinearity of the heat exchanges, we have developed an approach which is inspired by the Gappy POD and EIM methods. We use a multiphysics reduced basis that takes several contributions into account (temperatures, enthalpies, heat fluxes and humidities). The resulting reduced model is a hybrid model that couples some of the original physical equations to an artificial neural network. The reduction methodology has been validated on Renault vehicles. The reduced order models have been integrated into a vehicle system-level energetic simulation platform (GREEN) which models different thermics (engine, transmission, cooling system, battery, HVAC, refrigerant circuit, underhood) in order to perform thermal management studies which are of particular importance for electric and hybrid vehicles. The reduced order models have been validated on several scenarios (temperature control for thermal comfort, driving cycles, HVAC coupling) and have achieved CPU gains of up to 99% with average errors of 0.5 °C on temperatures and 0.6% on relative humidities
Maktoobi, Sheler. "Couplage diffractif pour réseaux de neurones optiques". Thesis, Bourgogne Franche-Comté, 2020. http://www.theses.fr/2020UBFCD019.
Pełny tekst źródłaPhotonic networks with high performance can be considered as substrates for future computing systems. In comparison with electronics, photonic systems have substantial privileges, for instance the possibility of a fully parallel implementation of networks. Recently, neural networks have moved into the center of attention of the photonic community. One of the most important requirements for parallel large-scale photonic networks is to realize the connectivities. Diffraction is considered as a method to process the connections between the nodes (coupling) in optical neural networks. In the current thesis, we evaluate the scalability of a diffractive coupling in more details as follow:First, we begin with a general introductions for artificial intelligence, machine learning, artificial neural network and photonic neural networks. To establish a working neural network, learning rules are an essential part to optimize a configuration for obtaining a low error from the system, hence learning rules are introduced (Chapter 1). We investigate the fundamental concepts of diffractive coupling in our spatio-temporal reservoir. In that case, theory of diffraction is explained. We use an analytical scheme to provide the limits for the size of diffractive networks which is a part of our photonic neural network (Chapter 2). The concepts of diffractive coupling are investigated experimentally by two different experiments to confirm the analytical limits and to obtain maximum number of nodes which can be coupled in the photonic network (Chapter 3). Numerical simulations for such an experimental setup is modeled in two different schemes to obtain the maximum size of network numerically, which approaches a surface of 100 mm2 (Chapter 4). Finally, the complete photonic neural network is demonstrated. We design a spatially extended reservoir for 900 nodes. Consequently, our system generalizes the prediction for the chaotic Mackey–Glass sequence (Chapter 5)
Ouali, Jamel. "Architecture intégrée flexible pour réseaux de neurones". Grenoble INPG, 1991. http://www.theses.fr/1991INPG0035.
Pełny tekst źródłaFernandez, Brillet Lucas. "Réseaux de neurones CNN pour la vision embarquée". Thesis, Université Grenoble Alpes, 2020. http://www.theses.fr/2020GRALM043.
Pełny tekst źródłaRecently, Convolutional Neural Networks have become the state-of-the-art soluion(SOA) to most computer vision problems. In order to achieve high accuracy rates, CNNs require a high parameter count, as well as a high number of operations. This greatly complicates the deployment of such solutions in embedded systems, which strive to reduce memory size. Indeed, while most embedded systems are typically in the range of a few KBytes of memory, CNN models from the SOA usually account for multiple MBytes, or even GBytes in model size. Throughout this thesis, multiple novel ideas allowing to ease this issue are proposed. This requires to jointly design the solution across three main axes: Application, Algorithm and Hardware.In this manuscript, the main levers allowing to tailor computational complexity of a generic CNN-based object detector are identified and studied. Since object detection requires scanning every possible location and scale across an image through a fixed-input CNN classifier, the number of operations quickly grows for high-resolution images. In order to perform object detection in an efficient way, the detection process is divided into two stages. The first stage involves a region proposal network which allows to trade-off recall for the number of operations required to perform the search, as well as the number of regions passed on to the next stage. Techniques such as bounding box regression also greatly help reduce the dimension of the search space. This in turn simplifies the second stage, since it allows to reduce the task’s complexity to the set of possible proposals. Therefore, parameter counts can greatly be reduced.Furthermore, CNNs also exhibit properties that confirm their over-dimensionment. This over-dimensionement is one of the key success factors of CNNs in practice, since it eases the optimization process by allowing a large set of equivalent solutions. However, this also greatly increases computational complexity, and therefore complicates deploying the inference stage of these algorithms on embedded systems. In order to ease this problem, we propose a CNN compression method which is based on Principal Component Analysis (PCA). PCA allows to find, for each layer of the network independently, a new representation of the set of learned filters by expressing them in a more appropriate PCA basis. This PCA basis is hierarchical, meaning that basis terms are ordered by importance, and by removing the least important basis terms, it is possible to optimally trade-off approximation error for parameter count. Through this method, it is possible to compress, for example, a ResNet-32 network by a factor of ×2 both in the number of parameters and operations with a loss of accuracy <2%. It is also shown that the proposed method is compatible with other SOA methods which exploit other CNN properties in order to reduce computational complexity, mainly pruning, winograd and quantization. Through this method, we have been able to reduce the size of a ResNet-110 from 6.88Mbytes to 370kbytes, i.e. a x19 memory gain with a 3.9 % accuracy loss.All this knowledge, is applied in order to achieve an efficient CNN-based solution for a consumer face detection scenario. The proposed solution consists of just 29.3kBytes model size. This is x65 smaller than other SOA CNN face detectors, while providing equal detection performance and lower number of operations. Our face detector is also compared to a more traditional Viola-Jones face detector, exhibiting approximately an order of magnitude faster computation, as well as the ability to scale to higher detection rates by slightly increasing computational complexity.Both networks are finally implemented in a custom embedded multiprocessor, verifying that theorical and measured gains from PCA are consistent. Furthermore, parallelizing the PCA compressed network over 8 PEs achieves a x11.68 speed-up with respect to the original network running on a single PE
Bigot, Pascal. "Utilisation des réseaux de neurones pour la télégestion des réseaux techniques urbains". Lyon 1, 1995. http://www.theses.fr/1995LYO10036.
Pełny tekst źródłaBénédic, Yohann. "Approche analytique pour l'optimisation de réseaux de neurones artificiels". Phd thesis, Université de Haute Alsace - Mulhouse, 2007. http://tel.archives-ouvertes.fr/tel-00605216.
Pełny tekst źródłaGatet, Laurent. "Intégration de Réseaux de Neurones pour la Télémétrie Laser". Phd thesis, Toulouse, INPT, 2007. http://oatao.univ-toulouse.fr/7595/1/gatet.pdf.
Pełny tekst źródłaChakik, Fadi El. "Maximum d'entropie et réseaux de neurones pour la classification". Grenoble INPG, 1998. http://www.theses.fr/1998INPG0091.
Pełny tekst źródłaBoné, Romuald. "Réseaux de neurones récurrents pour la prévision de séries temporelles". Tours, 2000. http://www.theses.fr/2000TOUR4003.
Pełny tekst źródłaOussar, Yacine. "Réseaux d'ondelettes et réseaux de neurones pour la modélisation statique et dynamique de processus". Phd thesis, Université Pierre et Marie Curie - Paris VI, 1998. http://pastel.archives-ouvertes.fr/pastel-00000677.
Pełny tekst źródłaChams, Ali. "Un environnement logiciel pour un simulateur de réseaux de neurones : smart". Grenoble INPG, 1992. http://www.theses.fr/1992INPG0103.
Pełny tekst źródłaStuner, Bruno. "Cohorte de réseaux de neurones récurrents pour la reconnaissance de l'écriture". Thesis, Normandie, 2018. http://www.theses.fr/2018NORMR024.
Pełny tekst źródłaState-of-the-art methods for handwriting recognition are based on LSTM recurrent neural networks (RNN) which achieve high performance recognition. In this thesis, we propose the lexicon verification and the cohort generation as two new building blocs to tackle the problem of handwriting recognition which are : i) the large vocabulary problem and the use of lexicon driven methods ii) the combination of multiple optical models iii) the need for large labeled dataset for training RNN. The lexicon verification is an alternative to the lexicon driven decoding process and can deal with lexicons of 3 millions words. The cohort generation is a method to get easily and quickly a large number of complementary recurrent neural networks extracted from a single training. From these two new techniques we build and propose a new cascade scheme for isolated word recognition, a new line level combination LV-ROVER and a new self-training strategy to train LSTM RNN for isolated handwritten words recognition. The proposed cascade combines thousands of LSTM RNN with lexicon verification and achieves state-of-the art word recognition performance on the Rimes and IAM datasets. The Lexicon Verified ROVER : LV-ROVER, has a reduce complexity compare to the original ROVER algorithm and combine hundreds of recognizers without language models while achieving state of the art for handwritten line text on the RIMES dataset. Our self-training strategy use both labeled and unlabeled data with the unlabeled data being self-labeled by its own lexicon verified predictions. The strategy enables self-training with a single BLSTM and show excellent results on the Rimes and Iam datasets
Weng, Qilong. "Stabilité pour des modèles de réseaux de neurones et de chimiotaxie". Thesis, Paris Sciences et Lettres (ComUE), 2017. http://www.theses.fr/2017PSLED026/document.
Pełny tekst źródłaThis thesis is aimed to study some biological models in neuronal network and chemotaxis with the spectral analysis method. In order to deal with the main concerning problems, such as the existence and uniqueness of the solutions and steady states as well as the asymptotic behaviors, the associated linear or linearized model is considered from the aspect of spectrum and semigroups in appropriate spaces then the nonlinear stability follows. More precisely, we start with a linear runs-and-tumbles equation in dimension d≥1 to establish the existence of a unique positive and normalized steady state and the exponential asymptotic stability in weighted L¹ space based on the Krein-Rutman theory together with some moment estimates from kinetic theory. Then, we consider time elapsed model under general assumptions on the firing rate and prove the uniqueness of the steady state and its nonlinear exponential stability in case without or with delay in the weak connectivity regime from the spectral analysis theory for semigroups. Finally, we study the model under weaker regularity assumption on the firing rate and the existence of the solution as well as the same exponential stability are established generally no matter taking delay into account or not and no matter in weak or strong connectivity regime
Latouche, Pierre. "Modèles de graphes aléatoires à structure cachée pour l'analyse des réseaux". Phd thesis, Université d'Evry-Val d'Essonne, 2010. http://tel.archives-ouvertes.fr/tel-00623088.
Pełny tekst źródłaKhalife, Sammy. "Graphes, géométrie et représentations pour le langage et les réseaux d'entités". Thesis, Institut polytechnique de Paris, 2020. http://www.theses.fr/2020IPPAX055.
Pełny tekst źródłaThe automated treatment of familiar objects, either natural or artifacts, always relies on a translation into entities manageable by computer programs. The choice of these abstract representations is always crucial for the efficiency of the treatments and receives the utmost attention from computer scientists and developers. However, another problem rises: the correspondence between the object to be treated and "its" representation is not necessarily one-to-one! Therefore, the ambiguous nature of certain discrete structures is problematic for their modeling as well as their processing and analysis with a program. Natural language, and in particular its textual representation, is an example. The subject of this thesis is to explore this question, which we approach using combinatorial and geometric methods. These methods allow us to address the problem of extracting information from large networks of entities and to construct representations useful for natural language processing.Firstly, we start by showing combinatorial properties of a family of graphs implicitly involved in sequential models. These properties essentially concern the inverse problem of finding a sequence representing a given graph. The resulting algorithms allow us to carry out an experimental comparison of different sequential models used in language modeling.Secondly, we consider an application for the problem of identifying named entities. Following a review of recent solutions, we propose a competitive method based on the comparison of knowledge graph structures which is less costly in annotating examples dedicated to the problem. We also establish an experimental analysis of the influence of entities from capital relations. This analysis suggests to broaden the framework for applying the identification of entities to knowledge bases of different natures. These solutions are used today in a software library in the banking sector.Then, we perform a geometric study of recently proposed representations of words, during which we discuss a geometric conjecture theoretically and experimentally. This study suggests that language analogies are difficult to transpose into geometric properties, and leads us to consider the paradigm of distance geometry in order to construct new representations.Finally, we propose a methodology based on the paradigm of distance geometry in order to build new representations of words or entities. We propose algorithms for solving this problem on some large scale instances, which allow us to build interpretable and competitive representations in performance for extrinsic tasks. More generally, we propose through this paradigm a new framework and research leadsfor the construction of representations in machine learning
Richard, Vincent. "Outils de synthèse pour les réseaux réflecteurs exploitant la cellule Phoenix et les réseaux de neurones". Thesis, Rennes, INSA, 2018. http://www.theses.fr/2018ISAR0004/document.
Pełny tekst źródłaIn collaboration with Thales Alenia Space and the French Space Agency (CNES), this PHD takes part in a very active international context on a new antenna: the reflectarrays (RA).Combining the advantages of conventional reflectors and those of networks, RA could eventually replace the currently used shaped reflectors. They consist of a primary source placed in front of a network of cells controlling the properties of the reflected electromagnetic field. Although many studies already focus on the characterization of cells, one of the issues is to carefully select them to achieve the final antenna: this is the synthesis step.An overview of different synthesis methods revealed the complexity to quickly obtain good performance simultaneously on the co- and cross-polarizations, for a wide frequency band and for the realization of shaped radiation pattern. The Phoenix cell is selected in this work for its good properties since it provides the entire phase range following a continuous cycle of geometries.Because one of the constraints in the design of RA is to maintain continuous geometry variations between two juxtaposed cells on the layout, a spherical representation tool made it possible to classify all the studied cells. It judiciously lists all the cells on a continuous, closed and periodic surface.A new step is reached with the design of behavioral models using Artificial Neural Networks (ANN). These models enable to a fast electromagnetic characterization of cells in terms of phase and amplitude of the direct and cross coefficients of the reflection matrix.The originality of the synthesis algorithm proposed in this work is the combined use of the spherical representation and a rapid cell characterization by ANN. A min / max optimization tool is used to improve the overall performance of the RA panel. It is then applied to a concrete case as part of a telecommunication mission
Albano, Alice. "Dynamique des graphes de terrain : analyse en temps intrinsèque". Thesis, Paris 6, 2014. http://www.theses.fr/2014PA066260/document.
Pełny tekst źródłaWe are surrounded by a multitude of interaction networks from different contexts. These networks can be modeled as graphs, called complex networks. They have a community structure, i.e. groups of nodes closely related to each other and less connected with the rest of the graph. An other phenomenon studied in complex networks in many contexts is diffusion. The spread of a disease is an example of diffusion. These phenomena are dynamic and depend on an important parameter, which is often little studied: the time scale in which they are observed. According to the chosen scale, the graph dynamics can vary significantly. In this thesis, we propose to study dynamic processes using a suitable time scale. We consider a notion of relative time which we call intrinsic time, opposed to "traditional" time, which we call extrinsic time. We first study diffusion phenomena using intrinsic time, and we compare our results with an extrinsic time scale. This allows us to highlight the fact that the same phenomenon observed at two different time scales can have a very different behavior. We then analyze the relevance of the use of intrinsic time scale for detecting dynamic communities. Comparing communities obtained according extrinsic and intrinsic scales shows that the intrinsic time scale allows a more significant detection than extrinsic time scale
Albano, Alice. "Dynamique des graphes de terrain : analyse en temps intrinsèque". Electronic Thesis or Diss., Paris 6, 2014. http://www.theses.fr/2014PA066260.
Pełny tekst źródłaWe are surrounded by a multitude of interaction networks from different contexts. These networks can be modeled as graphs, called complex networks. They have a community structure, i.e. groups of nodes closely related to each other and less connected with the rest of the graph. An other phenomenon studied in complex networks in many contexts is diffusion. The spread of a disease is an example of diffusion. These phenomena are dynamic and depend on an important parameter, which is often little studied: the time scale in which they are observed. According to the chosen scale, the graph dynamics can vary significantly. In this thesis, we propose to study dynamic processes using a suitable time scale. We consider a notion of relative time which we call intrinsic time, opposed to "traditional" time, which we call extrinsic time. We first study diffusion phenomena using intrinsic time, and we compare our results with an extrinsic time scale. This allows us to highlight the fact that the same phenomenon observed at two different time scales can have a very different behavior. We then analyze the relevance of the use of intrinsic time scale for detecting dynamic communities. Comparing communities obtained according extrinsic and intrinsic scales shows that the intrinsic time scale allows a more significant detection than extrinsic time scale
Elayyadi, Mohamed. "Equations aux dérivées partielles et réseaux de neurones pour le traitement d'images". Phd thesis, Université Joseph Fourier (Grenoble), 1997. http://tel.archives-ouvertes.fr/tel-00004940.
Pełny tekst źródłaRobitaille, Louis-Emile. "Réseaux de neurones pour l'apprentissage de la préférence en microscopie super-résolution". Master's thesis, Université Laval, 2021. http://hdl.handle.net/20.500.11794/68744.
Pełny tekst źródłaFor many years, fluorescent microscopy has been limited by diffraction. However, to study dynamic phenomena inside cells, a nanometric resolution is often necessary. To cope with this problem, an important development for fluorescent microscopy was the invention ofSTimulated-Emission-Depletion microscopy (STED) (Hell and Wichmann, 1994). If STEDachieves nanometric microscopy, it is also an extremely sophisticated technique that requires advanced knowledge across a wide range of domains, e.g. physics, chemistry and biology. With the goal of democratising the microscope, Durand et al. (2018) use the last development in artificial intelligence to automate STED parameterization with an optimisation loop. The objective aimed is to produce high-quality images while minimising photo bleaching and exposition time. The inability of measuring image quality and of choosing between compromise among objectives still forces an expert to stay behind the microscope. By automating the assessment of image quality and the selection of compromise, this master thesis intends to demonstrate the potential of neural networks for preference learning in life science.
Bourgeois, Yoann. "Les réseaux de neurones artificiels pour mesurer les risques économiques et financiers". Paris, EHESS, 2003. http://www.theses.fr/2003EHES0118.
Pełny tekst źródłaThe objective of this thesis is to provide complete methodologies to solve prediction and classification problems in economy and finance by using Artificial Neural networks. The plan of work shows that the thesisplays a great part in establishing in several ways a statistic methodology for neural networks. We proceed in four chapters. The first chapter describes supervised and unsupervised neural network methodology to modelize quantitative or qualitative variables. In the second chapter, we are interested by the bayesian approach for supervised neural networks and the developpement of a set of misspecification statistic tests for binary choice models. In chapter three, we show that multivariate supervised neural networks enable to take into account structural changes and the neural networks methodology is able to estimate some probabilities of exchange crisis. In chapter four, we develope a complete based neural network-GARCH model to manage a stocks portfolio. We introduce some terms as conditional returns or conditional risk for a stock or a portfolio. Next, we apply bayesian Self-Organizing Map in order to estimate the univariate probability density function of the DM/USD exchange rate
Rondel, Nadine. "Réseaux de neurones pour le traitement d'antenne et la commande référencée capteur". Brest, 1996. http://www.theses.fr/1996BRES2022.
Pełny tekst źródłaGautier, Eric. "Utilisation des réseaux de neurones artificiels pour la commande d'un véhicule autonome". Grenoble INPG, 1999. http://www.theses.fr/1999INPG0009.
Pełny tekst źródłaThe subject of this thesis covers both mobile robotic and artificial neural networks (ANN) fields. Our aim is to study solutions that connectionist techniques can bring to particular problems raised by the automatic control of a car-like vehicle. This report is composed of two main parts. The first of them processes fundamental aspects of mobile robot control and of the use of artificial neural networks for control of complex systems. This first study allows us to underline the different points where ANN can contribute in a control architecture providing a real autonomy to the vehicle while respecting the robustness and rapidity constraints induced by the utilisation of a robot of the size and the speed of a car. We propose in the second part of this report several controllers allowing gradual increase of the robot autonomy. First of all, we are interested in a simple task consisting only in enslaving the robot on a reference path given by a planner. Our approach enables a continuous adaptation of the system facing possible changes of the parameters of the robot or its environment. So as to allow the execution of manoeuvres without external orders, we also propose a methodology for the realisation of controllers based on external sensors of the vehicle. Our approach uses a model allying characteristics from both fuzzy logic and ANN. Finally we show how complex tasks can be realised using a sequence of several simple controllers. Our realisation of the selection system for these controllers, which uses a recurrent ANN, exhibits some characteristics of robustness and very fast reactions when faced to the external events that must be taken into account
Maghrebi, Fatine. "Modèles de réseaux de neurones pour la commande des carrefours à feux". Paris 1, 1994. http://www.theses.fr/1994PA010082.
Pełny tekst źródłaVidal, Martin. "Architecture systolique pour un algorithme basé sur les réseaux de neurones pour l'égalisation de canaux". Thèse, Université du Québec à Trois-Rivières, 1999. http://depot-e.uqtr.ca/3409/1/000662425.pdf.
Pełny tekst źródłaBernauer, Éric. "Les réseaux de neurones et l'aide au diagnostic : un modèle de neurones bouclés pour l'apprentissage de séquences temporelles". Toulouse 3, 1996. http://www.theses.fr/1996TOU30277.
Pełny tekst źródłaTremblay, Nicolas. "Réseaux et signal : des outils de traitement du signal pour l'analyse des réseaux". Thesis, Lyon, École normale supérieure, 2014. http://www.theses.fr/2014ENSL0938/document.
Pełny tekst źródłaThis thesis describes new tools specifically designed for the analysis of networks such as social, transportation, neuronal, protein, communication networks... These networks, along with the rapid expansion of electronic, IT and mobile technologies are increasingly monitored and measured. Adapted tools of analysis are therefore very much in demand, which need to be universal, powerful, and precise enough to be able to extract useful information from very different possibly large networks. To this end, a large community of researchers from various disciplines have concentrated their efforts on the analysis of graphs, well define mathematical tools modeling the interconnected structure of networks. Among all the considered directions of research, graph signal processing brings a new and promising vision : a signal is no longer defined on a regular n-dimensional topology, but on a particular topology defined by the graph. To apply these new ideas on the practical problems of network analysis paves the way to an analysis firmly rooted in signal processing theory. It is precisely this frontier between signal processing and network science that we explore throughout this thesis, as shown by two of its major contributions. Firstly, a multiscale version of community detection in networks is proposed, based on the recent definition of graph wavelets. Then, a network-adapted bootstrap method is introduced, that enables statistical estimation based on carefully designed graph resampling schemes
Fourure, Damien. "Réseaux de neurones convolutifs pour la segmentation sémantique et l'apprentissage d'invariants de couleur". Thesis, Lyon, 2017. http://www.theses.fr/2017LYSES056/document.
Pełny tekst źródłaComputer vision is an interdisciplinary field that investigates how computers can gain a high level of understanding from digital images or videos. In artificial intelligence, and more precisely in machine learning, the field in which this thesis is positioned,computer vision involves extracting characteristics from images and then generalizing concepts related to these characteristics. This field of research has become very popular in recent years, particularly thanks to the results of the convolutional neural networks that form the basis of so-called deep learning methods. Today, neural networks make it possible, among other things, to recognize different objects present in an image, to generate very realistic images or even to beat the champions at the Go game. Their performance is not limited to the image domain, since they are also used in other fields such as natural language processing (e. g. machine translation) or sound recognition. In this thesis, we study convolutional neural networks in order to develop specialized architectures and loss functions for low-level tasks (color constancy) as well as high-level tasks (semantic segmentation). Color constancy, is the ability of the human visual system to perceive constant colours for a surface despite changes in the spectrum of illumination (lighting change). In computer vision, the main approach consists in estimating the color of the illuminant and then suppressing its impact on the perceived color of objects. We approach the task of color constancy with the use of neural networks by developing a new architecture composed of a subsampling operator inspired by traditional methods. Our experience shows that our method makes it possible to obtain competitive performances with the state of the art. Nevertheless, our architecture requires a large amount of training data. In order to partially correct this problem and improve the training of neural networks, we present several techniques for artificial data augmentation. We are also making two contributions on a high-level issue : semantic segmentation. This task, which consists of assigning a semantic class to each pixel of an image, is a challenge in computer vision because of its complexity. On the one hand, it requires many examples of training that are costly to obtain. On the other hand, it requires the adaptation of traditional convolutional neural networks in order to obtain a so-called dense prediction, i. e., a prediction for each pixel present in the input image. To solve the difficulty of acquiring training data, we propose an approach that uses several databases annotated with different labels at the same time. To do this, we define a selective loss function that has the advantage of allowing the training of a convolutional neural network from data from multiple databases. We also developed self-context approach that captures the correlations between labels in different databases. Finally, we present our third contribution : a new convolutional neural network architecture called GridNet specialized for semantic segmentation. Unlike traditional networks, implemented with a single path from the input (image) to the output (prediction), our architecture is implemented as a 2D grid allowing several interconnected streams to operate at different resolutions. In order to exploit all the paths of the grid, we propose a technique inspired by dropout. In addition, we empirically demonstrate that our architecture generalize many of well-known stateof- the-art networks. We conclude with an analysis of the empirical results obtained with our architecture which, although trained from scratch, reveals very good performances, exceeding popular approaches often pre-trained
Lecomte, Jérémie. "Sélection de mode par réseaux de neurones pour le codeur audio AMR-WB+". Mémoire, Université de Sherbrooke, 2007. http://savoirs.usherbrooke.ca/handle/11143/1394.
Pełny tekst źródłaVasilache, Adriana. "Les réseaux de neurones pour la modélisation et la commande des procédés biotechnologiques". Toulouse, INSA, 2000. http://www.theses.fr/2000ISAT0050.
Pełny tekst źródłaIn this work we realize a study on the use of the neural nets for the modeling, classification and the control of fermentation processes. The black-box models (we consider a neural net like a black box model) are of great help for processes or phenomena modeling when analytical models cannot be deduced from physical considerations. Some of the advantages of the neural nets when compared to other black-box models are: they are universal approximators using a small number of parameters, their basis functions are adaptive, their repetitive structure permits an easy implementation both software and hardware and they have the property of implicit regularization. These, combined with the characteristics of the biological processes (which are non-linear, non-stationary processes whose dynamics isn’t entirely known), are the reason for which the neural nets are used for the modeling of such processes. We have thus used existing neural models and proposed new ones for the cases of lactic and alcoholic fermentations. We have presented two approaches for the characterization of the fermentation process dynamics: the modeling of the specific biomass growth rate, the most important dynamic parameter of a fermentation process and the global characterization of the process dynamics using a neural classifier. The two approaches have been tested in simulation and on real data for lactic or alcoholic fermentation processes. The use of a classifier of the process dynamics represents a potential tool for process supervision by means of detecting the changes in the process dynamics as well as an aid for the process modeling in the case of batch processes. The prediction of the biomass concentration has also been considered for a continuous fermentation process. The neural models have been tested in a predictive control strategy and compared with a similar strategy using adaptive modeling. The neural prediction has been an incontestable winner for the cases where the process dynamics changes in time. The last issue of our study has been the prediction of the respiratory quotient for a alcoholic fermentation for which we proposed a neural model. It has been proposed in view of a predictive control strategy for the maintenance of a certain regime (fermentative or oxidative)
Émirian, Frédéric. "Étude et conception d'une machine parallèle multi-modèles pour les réseaux de neurones". Toulouse, INPT, 1996. http://www.theses.fr/1996INPT091H.
Pełny tekst źródłaMouret, Jean-Baptiste. "Pressions sélectives multiples pour l'évolution de réseaux de neurones destinés à la robotique". Paris 6, 2008. http://www.theses.fr/2008PA066640.
Pełny tekst źródłaEvolutionary algoritms have been successfully used to generate controllers for many robots. However, they struggle to design complex artifacts when the fitness is unable to explicitly guide the process. In this thesis, we draw the hypothesis that these problems originate from the use of a single selection pressure, whereas living organisms are subject to many ones. We investigate here th euse of multiobjective evolutionary algorithms to create such multiple gradients in order to evolve neuro-controllers. We first describe how hypotheses about potential intermediate steps can be used by defining a multiobejctive optimization problem in which each objective corresponds to a sub-task. In the lack of any selection pressure, it is also possible to add an objective which encourages an efficient exploration of the neighborhood of current candidate solutions. We consider several possibilities to instantiate this concept for the evolution of neural networks and we conclude that maintaining the diversity of the behaviors, instead of the diversity of the genotype or the ohenotype, is an efficient way to override the deceptiveness of a fitness function. Last, we show that exaptations can be favored by applying a selection pressure on some modules of the generated neural-networks, possibly linked to genotypic modules. We tested these methods on the evolution of neural networks to compute a Boolean function and to control a light-seeking robot. They seem to be applicable to a wide range of evolutionary robotics problems, from complex locomotion to behavior control
Rey-Fabret, Isabelle. "Les réseaux de neurones pour la modélisation des écoulements d'effluents dans les pipelines". Paris, CNAM, 2005. http://www.theses.fr/2005CNAM0513.
Pełny tekst źródłaTACITE software gives a model of gas/liquid flow in the pipelines. It is composed of a thermodynamic module, an hydrodynamic module and a numerical scheme. It sometimes encounters difficulties because of the non derivability of the hydrodynamic function. The aim of the tesis is to assume the derivability of the hydrofynamic function. The aim of the thesis is to assume the derivability of this function by usin neural networks methodology. By analysing the problem, a multi-experts neural network is proposed. Its structure is based on the TACITE hydrodynamic module's one. The use of the HVS selection of varainles increases the model performances. Different integration tests in TACITE show that it is able to replace the TACITE hydrodynamic model, and can improve TACITE robustness. In conclusion, this thesis proposes a derivable model to globally reproduce the complex hydrodynamic phenomena generated by two phase flows
Cherif, Aymen. "Réseaux de neurones, SVM et approches locales pour la prévision de séries temporelles". Thesis, Tours, 2013. http://www.theses.fr/2013TOUR4003/document.
Pełny tekst źródłaTime series forecasting is a widely discussed issue for many years. Researchers from various disciplines have addressed it in several application areas : finance, medical, transportation, etc. In this thesis, we focused on machine learning methods : neural networks and SVM. We have also been interested in the meta-methods to push up the predictor performances, and more specifically the local models. In a divide and conquer strategy, the local models perform a clustering over the data sets before different predictors are affected into each obtained subset. We present in this thesis a new algorithm for recurrent neural networks to use them as local predictors. We also propose two novel clustering techniques suitable for local models. The first is based on Kohonen maps, and the second is based on binary trees
Ducoffe, Mélanie. "Active learning et visualisation des données d'apprentissage pour les réseaux de neurones profonds". Thesis, Université Côte d'Azur (ComUE), 2018. http://www.theses.fr/2018AZUR4115/document.
Pełny tekst źródłaOur work is presented in three separate parts which can be read independently. Firstly we propose three active learning heuristics that scale to deep neural networks: We scale query by committee, an ensemble active learning methods. We speed up the computation time by sampling a committee of deep networks by applying dropout on the trained model. Another direction was margin-based active learning. We propose to use an adversarial perturbation to measure the distance to the margin. We also establish theoretical bounds on the convergence of our Adversarial Active Learning strategy for linear classifiers. Some inherent properties of adversarial examples opens up promising opportunity to transfer active learning data from one network to another. We also derive an active learning heuristic that scales to both CNN and RNN by selecting the unlabeled data that minimize the variational free energy. Secondly, we focus our work on how to fasten the computation of Wasserstein distances. We propose to approximate Wasserstein distances using a Siamese architecture. From another point of view, we demonstrate the submodular properties of Wasserstein medoids and how to apply it in active learning. Eventually, we provide new visualization tools for explaining the predictions of CNN on a text. First, we hijack an active learning strategy to confront the relevance of the sentences selected with active learning to state-of-the-art phraseology techniques. These works help to understand the hierarchy of the linguistic knowledge acquired during the training of CNNs on NLP tasks. Secondly, we take advantage of deconvolution networks for image analysis to present a new perspective on text analysis to the linguistic community that we call Text Deconvolution Saliency
Sarr, Jean Michel Amath. "Étude de l’augmentation de données pour la robustesse des réseaux de neurones profonds". Electronic Thesis or Diss., Sorbonne université, 2023. http://www.theses.fr/2023SORUS072.
Pełny tekst źródłaIn this thesis, we considered the problem of the robustness of neural networks. That is, we have considered the case where the learning set and the deployment set are not independently and identically distributed from the same source. This hypothesis is called : the i.i.d hypothesis. Our main research axis has been data augmentation. Indeed, an extensive literature review and preliminary experiments showed us the regularization potential of data augmentation. Thus, as a first step, we sought to use data augmentation to make neural networks more robust to various synthetic and natural dataset shifts. A dataset shift being simply a violation of the i.i.d assumption. However, the results of this approach have been mixed. Indeed, we observed that in some cases the augmented data could lead to performance jumps on the deployment set. But this phenomenon did not occur every time. In some cases, the augmented data could even reduce performance on the deployment set. In our conclusion, we offer a granular explanation for this phenomenon. Better use of data augmentation toward neural network robustness is to generate stress tests to observe a model behavior when various shift occurs. Then, to use that information to estimate the error on the deployment set of interest even without labels, we call this deployment error estimation. Furthermore, we show that the use of independent data augmentation can improve deployment error estimation. We believe that this use of data augmentation will allow us to better quantify the reliability of neural networks when deployed on new unknown datasets
Quélavoine, Régis. "Etude de l'apprentissage et des structures des réseaux de neurones multicouches pour l'analyse de données". Avignon, 1997. http://www.theses.fr/1997AVIG0002.
Pełny tekst źródłaBétrouni, Mohamed. "Réseaux de neurones pour la projection plane de données multidimensionnelles et pour le suivi de procédés industriels". Lille 1, 1999. https://pepite-depot.univ-lille.fr/LIBRE/Th_Num/1999/50376-1999-21.pdf.
Pełny tekst źródłaGelly, Grégory. "Réseaux de neurones récurrents pour le traitement automatique de la parole". Thesis, Université Paris-Saclay (ComUE), 2017. http://www.theses.fr/2017SACLS295/document.
Pełny tekst źródłaAutomatic speech processing is an active field of research since the 1950s. Within this field the main area of research is automatic speech recognition but simpler tasks such as speech activity detection, language identification or speaker identification are also of great interest to the community. The most recent breakthrough in speech processing appeared around 2010 when speech recognition systems using deep neural networks drastically improved the state-of-the-art. Inspired by this gains and the work of Alex Graves on recurrent neural networks (RNN), we decided to explore the possibilities brought by these models on realistic data for two different tasks: speech activity detection and spoken language identification. In this work, we closely look at a specific model for the RNNs: the Long Short Term Memory (LSTM) which mitigates a lot of the difficulties that can arise when training an RNN. We augment this model and introduce optimization methods that lead to significant performance gains for speech activity detection and language identification. More specifically, we introduce a WER-like loss function to train a speech activity detection system so as to minimize the word error rate of a downstream speech recognition system. We also introduce two different methods to successfully train a multiclass classifier based on neural networks for tasks such as LID. The first one is based on a divide-and-conquer approach and the second one is based on an angular proximity loss function. Both yield performance gains but also speed up the training process
Borrel, Vincent. "Une approche intégrée pour la mobilité humaine dans les réseaux auto-organisables : causes et conséquences". Paris 6, 2008. http://www.theses.fr/2008PA066551.
Pełny tekst źródłaLécuyer, Fabrice. "Ordonner les nœuds pour passer à l'échelle sur les grands réseaux réels". Electronic Thesis or Diss., Sorbonne université, 2023. http://www.theses.fr/2023SORUS172.
Pełny tekst źródłaThis thesis focuses on using theoretical tools of computer science to improve algorithms in practice, specifically algorithms that process data in the form of graphs. A graph represents elements (nodes) and their interactions (edges). Computer scientists have designed theoretical algorithms for arbitrary graphs, such as finding shortest paths or identifying inter-connected nodes. However, real-world networks have specific properties that are unknown in advance due to the situations from which they arise. They can be very large, which presents a challenge for processing them in reasonable time. To help design scalable algorithms for real-world networks, we focus on the technique of node ordering, which consists in processing the nodes in a specific order that depends on local or global properties of the network. We provide a review on the different mechanisms and methods that have been used to design orderings across various application domains. Then, we present three contributions that use node orderings to make algorithms more efficient. First, we replicate a paper that designs an ordering to make cache systems more effective, which accelerates different graph algorithms. Second, we create new orderings that diminish the number of operations in an existing algorithm for triangle listing. Third, we use greedy algorithms with certain orderings to bound the size of a minimum vertex cover on a specific instance, which allows us to certify the quality of approximate values. These findings insist on scalability issues, time measurements, mathematical grounding and validation by experiments. Finally, we present a collaboration on network analysis that consists in describing the mobility of researchers within the space of knowledge
Toko, Worou Bio Mikaila. "Outils algorithmiques pour la détection des communautés dans les réseaux". Nice, 2012. http://www.theses.fr/2012NICE4097.
Pełny tekst źródłaThis thesis concerns the algorithmic aspects of the communities' detection in large graphs. The work can be used by a telecommunications operator whose graphs are associated to telephone calls and SMS or telecommunication networks. In this context, the detection of communities is used for the content recommendation, the analysis of customer data, the classification of Web pages, the detection of Web spamming, marketing activities and others. This thesis is organized around two major parts. In the first part, we introduce the field of detection of communities. Indeed this problem has been studied with different points of view during the last years. The main methods and applications are presented in this descriptive part. In the second part, we present our contribution to the problema. Our contribution consists of two main topics. First, we introduce a new quality function, the fractional arboricity which is more adapted to the problem of detecting communities in social networks. Then, we present a fast and performance guaranteed algorithm to approximate the optimal fractional arboricity and identifies the communities in question. Second, we study the detection of communities by optimizing the modularity, the most used quality function for communities’ detection. We rewrite this function, and then, find new interpretations of the modularity and also links between the modularity and others cut functions. Finally, we propose two heuristics to approximate the optimization of the modularity. The first is an algorithm that approximates the modularity by using the Fiedler vector of the Laplacian matrix of the graph. The second algorithm is a fast heuristic based on the representation of physical interaction of nodes in a metric space. With this representation, we define an attraction/ repulsion mechanism between the vertices and then we obtain clusters in communities. Finally, we combine the optimization of the fractional arboricity and the optimization of the modularity into one communities’ detection tool
Hollocou, Alexandre. "Nouvelles approches pour le partitionnement de grands graphes". Thesis, Paris Sciences et Lettres (ComUE), 2018. http://www.theses.fr/2018PSLEE063.
Pełny tekst źródłaGraphs 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
Kara, Reda. "Une Approche modulaire du réseau de neurones CMAC pour la commande d'un système robot-vision". Mulhouse, 2002. http://www.theses.fr/2002MULH0704.
Pełny tekst źródłaThe work of this thesis investigates artificial neural networks capabilities to estimate robotic functions, and their performances as controllers. We propose an adaptive visual servoing scheme based on the CMAC ("Cerebellar Model Articulation Controller") network. The CMAC network is thus well suited for robot control but in practice there are two drawbacks: its output is "discrete" and its precision depends on its size. Thus, we have developed two modular neural : the HCMAC ("Hierarchical CMAC") and the AL_CMAC ("Adaptive Linear CMAC"). These two networks are a combination of networks of small size. The efficiency of the HCMAC and AL_CMAC neuro-controller is validated through visual servoing experiments with a three degrees of freedom robot arm and with a two camera vision system. Visual servoing experiments consist in positioning tasks and in tracking mobile objects. The performances are compared to other neuro-controllers like CMAC and SSOM ("Supervised Self-Organizing Maps") networks
Rochel, Olivier. "Une approche événementielle pour la modélisation et la simulation de réseaux de neurones impulsionnels". Nancy 1, 2004. http://www.theses.fr/2004NAN10004.
Pełny tekst źródłaAt present, there exists no generic modeling and simulation framework that addresses the study of large spiking neural networks. In the existing models, the impulses are generally associated with discontinuities in the otherwise continuous dynamics of the neurons. This raises modeling and practical implementation issues. We propose an novel approach based on the discrete-event system abstraction, grounded on the DEVS formalism, that can be used to represent a large class of spiking neurons and permits the modeling of large networks. A simulation engine has been developed on top of this formalism. It is based on an efficient event-driven algorithm and has been evaluated on sequential as well as parallel machines. We have tested our approach within a multi-disciplinary project on olfactory perception
Bouaziz, Mohamed. "Réseaux de neurones récurrents pour la classification de séquences dans des flux audiovisuels parallèles". Thesis, Avignon, 2017. http://www.theses.fr/2017AVIG0224/document.
Pełny tekst źródłaIn the same way as TV channels, data streams are represented as a sequence of successive events that can exhibit chronological relations (e.g. a series of programs, scenes, etc.). For a targeted channel, broadcast programming follows the rules defined by the channel itself, but can also be affected by the programming of competing ones. In such conditions, event sequences of parallel streams could provide additional knowledge about the events of a particular stream. In the sphere of machine learning, various methods that are suited for processing sequential data have been proposed. Long Short-Term Memory (LSTM) Recurrent Neural Networks have proven its worth in many applications dealing with this type of data. Nevertheless, these approaches are designed to handle only a single input sequence at a time. The main contribution of this thesis is about developing approaches that jointly process sequential data derived from multiple parallel streams. The application task of our work, carried out in collaboration with the computer science laboratory of Avignon (LIA) and the EDD company, seeks to predict the genre of a telecast. This prediction can be based on the histories of previous telecast genres in the same channel but also on those belonging to other parallel channels. We propose a telecast genre taxonomy adapted to such automatic processes as well as a dataset containing the parallel history sequences of 4 French TV channels. Two original methods are proposed in this work in order to take into account parallel stream sequences. The first one, namely the Parallel LSTM (PLSTM) architecture, is an extension of the LSTM model. PLSTM simultaneously processes each sequence in a separate recurrent layer and sums the outputs of each of these layers to produce the final output. The second approach, called MSE-SVM, takes advantage of both LSTM and Support Vector Machines (SVM) methods. Firstly, latent feature vectors are independently generated for each input stream, using the output event of the main one. These new representations are then merged and fed to an SVM algorithm. The PLSTM and MSE-SVM approaches proved their ability to integrate parallel sequences by outperforming, respectively, the LSTM and SVM models that only take into account the sequences of the main stream. The two proposed approaches take profit of the information contained in long sequences. However, they have difficulties to deal with short ones. Though MSE-SVM generally outperforms the PLSTM approach, the problem experienced with short sequences is more pronounced for MSE-SVM. Finally, we propose to extend this approach by feeding additional information related to each event in the input sequences (e.g. the weekday of a telecast). This extension, named AMSE-SVM, has a remarkably better behavior with short sequences without affecting the performance when processing long ones