Дисертації з теми "Apprentissage par renforcement non supervisé"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся з топ-50 дисертацій для дослідження на тему "Apprentissage par renforcement non supervisé".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Переглядайте дисертації для різних дисциплін та оформлюйте правильно вашу бібліографію.
Tarbouriech, Jean. "Goal-oriented exploration for reinforcement learning." Thesis, Université de Lille (2022-....), 2022. http://www.theses.fr/2022ULILB014.
Learning to reach goals is a competence of high practical relevance to acquire for intelligent agents. For instance, this encompasses many navigation tasks ("go to target X"), robotic manipulation ("attain position Y of the robotic arm"), or game-playing scenarios ("win the game by fulfilling objective Z"). As a living being interacting with the world, I am constantly driven by goals to reach, varying in scope and difficulty.Reinforcement Learning (RL) holds the promise to frame and learn goal-oriented behavior. Goals can be modeled as specific configurations of the environment that must be attained via sequential interaction and exploration of the unknown environment. Although various deep RL algorithms have been proposed for goal-oriented RL, existing methods often lack principled understanding, sample efficiency and general-purpose effectiveness. In fact, very limited theoretical analysis of goal-oriented RL was available, even in the basic scenario of finitely many states and actions.We first focus on a supervised scenario of goal-oriented RL, where a goal state to be reached in minimum total expected cost is provided as part of the problem definition. After formalizing the online learning problem in this setting often known as Stochastic Shortest Path (SSP), we introduce two no-regret algorithms (one is the first available in the literature, the other attains nearly optimal guarantees).Beyond training our RL agent to solve only one task, we then aspire that it learns to autonomously solve a wide variety of tasks, in the absence of any reward supervision. In this challenging unsupervised RL scenario, we advocate to "Set Your Own Goals" (SYOG), which suggests the agent to learn the ability to intrinsically select and reach its own goal states. We derive finite-time guarantees of this popular heuristic in various settings, each with its specific learning objective and technical challenges. As an illustration, we propose a rigorous analysis of the algorithmic principle of targeting "uncertain" goals which we also anchor in deep RL.The main focus and contribution of this thesis are to instigate a principled analysis of goal-oriented exploration in RL, both in the supervised and unsupervised scenarios. We hope that it helps suggest promising research directions to improve the interpretability and sample efficiency of goal-oriented RL algorithms in practical applications
Debard, Quentin. "Automatic learning of next generation human-computer interactions." Thesis, Lyon, 2020. http://www.theses.fr/2020LYSEI036.
Artificial Intelligence (AI) and Human-Computer Interactions (HCIs) are two research fields with relatively few common work. HCI specialists usually design the way we interact with devices directly from observations and measures of human feedback, manually optimizing the user interface to better fit users’ expectations. This process is hard to optimize: ergonomy, intuitivity and ease of use are key features in a User Interface (UI) that are too complex to be simply modelled from interaction data. This drastically restrains the possible uses of Machine Learning (ML) in this design process. Currently, ML in HCI is mostly applied to gesture recognition and automatic display, e.g. advertisement or item suggestion. It is also used to fine tune an existing UI to better optimize it, but as of now it does not participate in designing new ways to interact with computers. Our main focus in this thesis is to use ML to develop new design strategies for overall better UIs. We want to use ML to build intelligent – understand precise, intuitive and adaptive – user interfaces using minimal handcrafting. We propose a novel approach to UI design: instead of letting the user adapt to the interface, we want the interface and the user to adapt mutually to each other. The goal is to reduce human bias in protocol definition while building co-adaptive interfaces able to further fit individual preferences. In order to do so, we will put to use the different mechanisms available in ML to automatically learn behaviors, build representations and take decisions. We will be experimenting on touch interfaces, as these interfaces are vastly used and can provide easily interpretable problems. The very first part of our work will focus on processing touch data and use supervised learning to build accurate classifiers of touch gestures. The second part will detail how Reinforcement Learning (RL) can be used to model and learn interaction protocols given user actions. Lastly, we will combine these RL models with unsupervised learning to build a setup allowing for the design of new interaction protocols without the need for real user data
Buhot, Arnaud. "Etude de propriétés d'apprentissage supervisé et non supervisé par des méthodes de Physique Statistique." Phd thesis, Université Joseph Fourier (Grenoble), 1999. http://tel.archives-ouvertes.fr/tel-00001642.
Chen, Hao. "Vers la ré-identification de personnes non-supervisée." Thesis, Université Côte d'Azur, 2022. http://www.theses.fr/2022COAZ4014.
As a core component of intelligent video surveillance systems, person re-identification (ReID) targets at retrieving a person of interest across non-overlapping cameras. Despite significant improvements in supervised ReID, cumbersome annotation process makes it less scalable in real-world deployments. Moreover, as appearance representations can be affected by noisy factors, such as illumination level and camera properties, between different domains, person ReID models suffer a large performance drop in the presence of domain gaps. We are particularly interested in designing algorithms that can adapt a person ReID model to a target domain without human supervision. In such context, we mainly focus on designing unsupervised domain adaptation and unsupervised representation learning methods for person ReID.In this thesis, we first explore how to build robust representations by combining both global and local features under the supervised condition. Then, towards an unsupervised domain adaptive ReID system, we propose three unsupervised methods for person ReID, including 1) teacher-student knowledge distillation with asymmetric network structures for feature diversity encouragement, 2) joint generative and contrastive learning framework that generates augmented views with a generative adversarial network for contrastive learning, and 3) exploring inter-instance relations and designing relation-aware loss functions for better contrastive learning based person ReID.Our methods have been extensively evaluated on main-stream ReID datasets, such as Market-1501, DukeMTMC-reID and MSMT17. The proposed methods significantly outperform previous methods on the ReID datasets, significantly pushing person ReID to real-world deployments
Dutech, Alain. "Apprentissage par Renforcement : Au delà des Processus Décisionnels de Markov (Vers la cognition incarnée)." Habilitation à diriger des recherches, Université Nancy II, 2010. http://tel.archives-ouvertes.fr/tel-00549108.
Lefort, Mathieu. "Apprentissage spatial de corrélations multimodales par des mécanismes d'inspiration corticale." Phd thesis, Université Nancy II, 2012. http://tel.archives-ouvertes.fr/tel-00756687.
Peyrache, Jean-Philippe. "Nouvelles approches itératives avec garanties théoriques pour l'adaptation de domaine non supervisée." Thesis, Saint-Etienne, 2014. http://www.theses.fr/2014STET4023/document.
During the past few years, an increasing interest for Machine Learning has been encountered, in various domains like image recognition or medical data analysis. However, a limitation of the classical PAC framework has recently been highlighted. It led to the emergence of a new research axis: Domain Adaptation (DA), in which learning data are considered as coming from a distribution (the source one) different from the one (the target one) from which are generated test data. The first theoretical works concluded that a good performance on the target domain can be obtained by minimizing in the same time the source error and a divergence term between the two distributions. Three main categories of approaches are derived from this idea : by reweighting, by reprojection and by self-labeling. In this thesis work, we propose two contributions. The first one is a reprojection approach based on boosting theory and designed for numerical data. It offers interesting theoretical guarantees and also seems able to obtain good generalization performances. Our second contribution consists first in a framework filling the gap of the lack of theoretical results for self-labeling methods by introducing necessary conditions ensuring the good behavior of this kind of algorithm. On the other hand, we propose in this framework a new approach, using the theory of (epsilon, gamma, tau)- good similarity functions to go around the limitations due to the use of kernel theory in the specific context of structured data
De, La Bourdonnaye François. "Learning sensori-motor mappings using little knowledge : application to manipulation robotics." Thesis, Université Clermont Auvergne (2017-2020), 2018. http://www.theses.fr/2018CLFAC037/document.
The thesis is focused on learning a complex manipulation robotics task using little knowledge. More precisely, the concerned task consists in reaching an object with a serial arm and the objective is to learn it without camera calibration parameters, forward kinematics, handcrafted features, or expert demonstrations. Deep reinforcement learning algorithms suit well to this objective. Indeed, reinforcement learning allows to learn sensori-motor mappings while dispensing with dynamics. Besides, deep learning allows to dispense with handcrafted features for the state spacerepresentation. However, it is difficult to specify the objectives of the learned task without requiring human supervision. Some solutions imply expert demonstrations or shaping rewards to guiderobots towards its objective. The latter is generally computed using forward kinematics and handcrafted visual modules. Another class of solutions consists in decomposing the complex task. Learning from easy missions can be used, but this requires the knowledge of a goal state. Decomposing the whole complex into simpler sub tasks can also be utilized (hierarchical learning) but does notnecessarily imply a lack of human supervision. Alternate approaches which use several agents in parallel to increase the probability of success can be used but are costly. In our approach,we decompose the whole reaching task into three simpler sub tasks while taking inspiration from the human behavior. Indeed, humans first look at an object before reaching it. The first learned task is an object fixation task which is aimed at localizing the object in the 3D space. This is learned using deep reinforcement learning and a weakly supervised reward function. The second task consists in learning jointly end-effector binocular fixations and a hand-eye coordination function. This is also learned using a similar set-up and is aimed at localizing the end-effector in the 3D space. The third task uses the two prior learned skills to learn to reach an object and uses the same requirements as the two prior tasks: it hardly requires supervision. In addition, without using additional priors, an object reachability predictor is learned in parallel. The main contribution of this thesis is the learning of a complex robotic task with weak supervision
Aklil, Nassim. "Apprentissage actif sous contrainte de budget en robotique et en neurosciences computationnelles. Localisation robotique et modélisation comportementale en environnement non stationnaire." Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066225/document.
Decision-making is a highly researched field in science, be it in neuroscience to understand the processes underlying animal decision-making, or in robotics to model efficient and rapid decision-making processes in real environments. In neuroscience, this problem is resolved online with sequential decision-making models based on reinforcement learning. In robotics, the primary objective is efficiency, in order to be deployed in real environments. However, in robotics what can be called the budget and which concerns the limitations inherent to the hardware, such as computation times, limited actions available to the robot or the lifetime of the robot battery, are often not taken into account at the present time. We propose in this thesis to introduce the notion of budget as an explicit constraint in the robotic learning processes applied to a localization task by implementing a model based on work developed in statistical learning that processes data under explicit constraints, limiting the input of data or imposing a more explicit time constraint. In order to discuss an online functioning of this type of budgeted learning algorithms, we also discuss some possible inspirations that could be taken on the side of computational neuroscience. In this context, the alternation between information retrieval for location and the decision to move for a robot may be indirectly linked to the notion of exploration-exploitation compromise. We present our contribution to the modeling of this compromise in animals in a non-stationary task involving different levels of uncertainty, and we make the link with the methods of multi-armed bandits
Maes, Francis. "Learning in Markov decision processes for structured prediction : applications to sequence labeling, tree transformation and learning for search." Paris 6, 2009. http://www.theses.fr/2009PA066500.
Muller, Jean-Denis. "La perception structurante : apprentissage non monotone de fonctions visuelles par croissance et maturation de structures neuromimétiques." Toulouse, ENSAE, 1993. http://www.theses.fr/1993ESAE0030.
Wacongne, Catherine. "Traitements conscient et non-conscient des régularités temporelles : Modélisation et neuroimagerie." Thesis, Paris 6, 2014. http://www.theses.fr/2014PA066290/document.
What is going to happen next? Natural stimuli tend to follow each other in a reproducible way. Multiple fields of neuroscience and psychology bring evidence that human’s brain and behavior are sensitive to the temporal structure of stimuli and are able to exploit them in multiple ways: to make appropriate decisions, encode efficiently information, react faster to predictable stimuli or orient attention towards surprising ones… Multiple brain areas show sensitivity to the temporal structure of events. However, all areas do not seem to be sensitive to the same kind of temporal regularities. Conscious access to the stimuli seems to play a key role in some of these dissociations and better understanding this role could improve the current diagnostic tools for non-communicative patients. This thesis explores the hierarchical organization of the processing of temporal regularities and the computational properties of conscious and unconscious levels of processing by combining a modeling approach with neuroimaging experiments using magnetoencephalography and electroencephalography (MEEG). First, a plausible neuronal model based on predictive coding principles reproduces the main properties of the preattentive processing of pure tones in the auditory cortex indexed by the evoked potential mismatch negativity (MMN). Second, a MEEG experiment provides evidence for a hierarchical organization of multiple predictive processes in the auditory cortex. Finally, a second model explores the new computational properties and constraints associated to the access of stimuli to a conscious space with a working memory able to maintain information for an arbitrary time but with limited capacity
Laumônier, Julien. "Méthodes d'apprentissage de la coordination multiagent : application au transport intelligent." Doctoral thesis, Université Laval, 2008. http://hdl.handle.net/20.500.11794/20000.
Monnet, Jean-matthieu. "Caractérisation des forêts de montagne par scanner laser aéroporté : estimation de paramètres de peuplement par régression SVM et apprentissage non supervisé pour la détection de sommets." Phd thesis, Université de Grenoble, 2011. http://tel.archives-ouvertes.fr/tel-00652698.
Monnet, Jean-Matthieu. "Caractérisation des forêts de montagne par scanner laser aéroporté : estimation de paramètres de peuplement par régression SVM et apprentissage non supervisé pour la détection de sommets." Thesis, Grenoble, 2011. http://www.theses.fr/2011GRENT056/document.
Numerous studies have shown the potential of airborne laser scanningfor the mapping of forest resources. However, the application of thisremote sensing technique to complex forests encountered in mountainousareas requires further investigation. In this thesis, the two mainmethods used to derive forest information are tested with airbornelaser scanning data acquired in the French Alps, and adapted to theconstraints of mountainous environments. In particular,a framework for unsupervised training of treetop detection isproposed, and the performance of support vector regression combinedwith dimension reduction for forest stand parameters estimation isevaluated
Najar, Anis. "Shaping robot behaviour with unlabeled human instructions." Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066152.
Most of current interactive learning systems rely on predefined protocols that constrain the interaction with the user. Relaxing the constraints of interaction protocols can therefore improve the usability of these systems.This thesis tackles the question of interpreting human instructions, in order to relax the constraints about predetermining their meanings. We propose a framework that enables a human teacher to shape a robot behaviour, by interactively providing it with unlabeled instructions. Our approach consists in grounding the meaning of instruction signals in the task learning process, and using them simultaneously for guiding the latter. This approach has a two-fold advantage. First, it provides more freedom to the teacher in choosing his preferred signals. Second, it reduces the required engineering efforts, by removing the necessity to encode the meaning of each instruction signal. We implement our framework as a modular architecture, named TICS, that offers the possibility to combine different information sources: a predefined reward function, evaluative feedback and unlabeled instructions. This allows for more flexibility in the teaching process, by enabling the teacher to switch between different learning modes. Particularly, we propose several methods for interpreting instructions, and a new method for combining evaluative feedback with a predefined reward function. We evaluate our framework through a series of experiments, performed both in simulation and with real robots. The experimental results demonstrate the effectiveness of our framework in accelerating the task learning process, and in reducing the number of required interactions with the teacher
Sahasrabudhe, Mihir. "Unsupervised and weakly supervised deep learning methods for computer vision and medical imaging." Thesis, université Paris-Saclay, 2020. http://www.theses.fr/2020UPASC010.
The first two contributions of this thesis (Chapter 2 and 3) are models for unsupervised 2D alignment and learning 3D object surfaces, called Deforming Autoencoders (DAE) and Lifting Autoencoders (LAE). These models are capable of identifying canonical space in order to represent different object properties, for example, appearance in a canonical space, deformation associated with this appearance that maps it to the image space, and for human faces, a 3D model for a face, its facial expression, and the angle of the camera. We further illustrate applications of models to other domains_ alignment of lung MRI images in medical image analysis, and alignment of satellite images for remote sensing imagery. In Chapter 4, we concentrate on a problem in medical image analysis_ diagnosis of lymphocytosis. We propose a convolutional network to encode images of blood smears obtained from a patient, followed by an aggregation operation to gather information from all images in order to represent them in one feature vector which is used to determine the diagnosis. Our results show that the performance of the proposed models is at-par with biologists and can therefore augment their diagnosis
Dubois, Amaury. "Optimisation et apprentissage de modèles biologiques : application à lirrigation [sic l'irrigation] de pomme de terre." Thesis, Littoral, 2020. http://www.theses.fr/2020DUNK0560.
The subject of this PhD concerns one of the LISIC themes : modelling and simulation of complex systems, as well as optimization and automatic learning for agronomy. The objectives of the thesis are to answer the questions of irrigation management of the potato crop and the development of decision support tools for farmers. The choice of this crop is motivated by its important share in the Haut-de-France region. The manuscript is divided into 3 parts. The first part deals with continuous multimodal optimization in a black box context. This is followed by a presentation of a methodology for the automatic calibration of biological model parameters through reformulation into a black box multimodal optimization problem. The relevance of the use of inverse analysis as a methodology for automatic parameterisation of large models in then demonstrated. The second part presents 2 new algorithms, UCB Random with Decreasing Step-size and UCT Random with Decreasing Step-size. Thes algorithms are designed for continuous multimodal black-box optimization whose choice of the position of the initial local search is assisted by a reinforcement learning algorithms. The results show that these algorithms have better performance than (Quasi) Random with Decreasing Step-size algorithms. Finally, the last part focuses on machine learning principles and methods. A reformulation of the problem of predicting soil water content at one-week intervals into a supervised learning problem has enabled the development of a new decision support tool to respond to the problem of crop management
Hedjazi, Lyamine. "Outil d'aide au diagnostic du cancer à partir d'extraction d'informations issues de bases de données et d'analyses par biopuces." Phd thesis, Université Paul Sabatier - Toulouse III, 2011. http://tel.archives-ouvertes.fr/tel-00657959.
Lucas, Thomas. "Modèles génératifs profonds : sur-généralisation et abandon de mode." Thesis, Université Grenoble Alpes, 2020. http://www.theses.fr/2020GRALM049.
This dissertation explores the topic of generative modelling of natural images,which is the task of fitting a data generating distribution.Such models can be used to generate artificial data resembling the true data, or to compress images.Latent variable models, which are at the core of our contributions, seek to capture the main factors of variations of an image into a variable that can be manipulated.In particular we build on two successful latent variable generative models, the generative adversarial network (GAN) and Variational autoencoder (VAE) models.Recently GANs significantly improved the quality of images generated by deep models, obtaining very compelling samples.Unfortunately these models struggle to capture all the modes of the original distribution, ie they do not cover the full variability of the dataset.Conversely, likelihood based models such as VAEs typically cover the full variety of the data well and provide an objective measure of coverage.However these models produce samples of inferior visual quality that are more easily distinguished from real ones.The work presented in this thesis strives for the best of both worlds: to obtain compelling samples while modelling the full support of the distribution.To achieve that, we focus on i) the optimisation problems used and ii) practical model limitations that hinder performance.The first contribution of this manuscript is a deep generative model that encodes global image structure into latent variables, built on the VAE, and autoregressively models low level detail.We propose a training procedure relying on an auxiliary loss function to control what information is captured by the latent variables and what information is left to an autoregressive decoder.Unlike previous approaches to such hybrid models, ours does not need to restrict the capacity of the autoregressive decoder to prevent degenerate models that ignore the latent variables.The second contribution builds on the standard GAN model, which trains a discriminator network to provide feedback to a generative network.The discriminator usually assesses the quality of individual samples, which makes it hard to evaluate the variability of the data.Instead we propose to feed the discriminator with emph{batches} that mix both true and fake samples, and train it to predict the ratio of true samples in the batch.These batches work as approximations of the distribution of generated images and allows the discriminator to approximate distributional statistics.We introduce an architecture that is well suited to solve this problem efficiently,and show experimentally that our approach reduces mode collapse in GANs on two synthetic datasets, and obtains good results on the CIFAR10 and CelebA datasets.The mutual shortcomings of VAEs and GANs can in principle be addressed by training hybrid models that use both types of objective.In our third contribution, we show that usual parametric assumptions made in VAEs induce a conflict between them, leading to lackluster performance of hybrid models.We propose a solution based on deep invertible transformations, that trains a feature space in which usual assumptions can be made without harm.Our approach provides likelihood computations in image space while being able to take advantage of adversarial training.It obtains GAN-like samples that are competitive with fully adversarial models while improving likelihood scores over existing hybrid models at the time of publication, which is a significant advancement
Chahla, Charbel. "Non-linear feature extraction for object re-identification in cameras networks." Thesis, Troyes, 2017. http://www.theses.fr/2017TROY0023.
Replicating the visual system that the brain uses to process the information is an area of substantial interest. This thesis is situated in the context of a fully automated system capable of analyzing facial features when the target is near the cameras, and tracking his identity when his facial features are no more traceable. The first part of this thesis is devoted to face pose estimation procedures to be used in face recognition scenarios. We proposed a new label-sensitive embedding based on a sparse representation called Sparse Label sensitive Locality Preserving Projections. In an uncontrolled environment observed by cameras from an unknown distance, person re-identification relying upon conventional biometrics such as face recognition is not feasible. Instead, visual features based on the appearance of people can be exploited more reliably. In this context, we propose a new embedding scheme for single-shot person re-identification under non overlapping target cameras. Each person is described as a vector of kernel similarities to a collection of prototype person images. The robustness of the algorithm is improved by proposing the Color Categorization procedure. In the last part of this thesis, we propose a Siamese architecture of two Convolutional Neural Networks (CNN), with each CNN reduced to only eleven layers. This architecture allows a machine to be fed directly with raw data and to automatically discover the representations needed for classification
Ho, Vinh Thanh. "Techniques avancées d'apprentissage automatique basées sur la programmation DC et DCA." Thesis, Université de Lorraine, 2017. http://www.theses.fr/2017LORR0289/document.
In this dissertation, we develop some advanced machine learning techniques in the framework of online learning and reinforcement learning (RL). The backbones of our approaches are DC (Difference of Convex functions) programming and DCA (DC Algorithm), and their online version that are best known as powerful nonsmooth, nonconvex optimization tools. This dissertation is composed of two parts: the first part studies some online machine learning techniques and the second part concerns RL in both batch and online modes. The first part includes two chapters corresponding to online classification (Chapter 2) and prediction with expert advice (Chapter 3). These two chapters mention a unified DC approximation approach to different online learning algorithms where the observed objective functions are 0-1 loss functions. We thoroughly study how to develop efficient online DCA algorithms in terms of theoretical and computational aspects. The second part consists of four chapters (Chapters 4, 5, 6, 7). After a brief introduction of RL and its related works in Chapter 4, Chapter 5 aims to provide effective RL techniques in batch mode based on DC programming and DCA. In particular, we first consider four different DC optimization formulations for which corresponding attractive DCA-based algorithms are developed, then carefully address the key issues of DCA, and finally, show the computational efficiency of these algorithms through various experiments. Continuing this study, in Chapter 6 we develop DCA-based RL techniques in online mode and propose their alternating versions. As an application, we tackle the stochastic shortest path (SSP) problem in Chapter 7. Especially, a particular class of SSP problems can be reformulated in two directions as a cardinality minimization formulation and an RL formulation. Firstly, the cardinality formulation involves the zero-norm in objective and the binary variables. We propose a DCA-based algorithm by exploiting a DC approximation approach for the zero-norm and an exact penalty technique for the binary variables. Secondly, we make use of the aforementioned DCA-based batch RL algorithm. All proposed algorithms are tested on some artificial road networks
Kannan, Hariprasad. "Quelques applications de l’optimisation numérique aux problèmes d’inférence et d’apprentissage." Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLC067/document.
Numerical optimization and machine learning have had a fruitful relationship, from the perspective of both theory and application. In this thesis, we present an application oriented take on some inference and learning problems. Linear programming relaxations are central to maximum a posteriori (MAP) inference in discrete Markov Random Fields (MRFs). Especially, inference in higher-order MRFs presents challenges in terms of efficiency, scalability and solution quality. In this thesis, we study the benefit of using Newton methods to efficiently optimize the Lagrangian dual of a smooth version of the problem. We investigate their ability to achieve superior convergence behavior and to better handle the ill-conditioned nature of the formulation, as compared to first order methods. We show that it is indeed possible to obtain an efficient trust region Newton method, which uses the true Hessian, for a broad range of MAP inference problems. Given the specific opportunities and challenges in the MAP inference formulation, we present details concerning (i) efficient computation of the Hessian and Hessian-vector products, (ii) a strategy to damp the Newton step that aids efficient and correct optimization, (iii) steps to improve the efficiency of the conjugate gradient method through a truncation rule and a pre-conditioner. We also demonstrate through numerical experiments how a quasi-Newton method could be a good choice for MAP inference in large graphs. MAP inference based on a smooth formulation, could greatly benefit from efficient sum-product computation, which is required for computing the gradient and the Hessian. We show a way to perform sum-product computation for trees with sparse clique potentials. This result could be readily used by other algorithms, also. We show results demonstrating the usefulness of our approach using higher-order MRFs. Then, we discuss potential research topics regarding tightening the LP relaxation and parallel algorithms for MAP inference.Unsupervised learning is an important topic in machine learning and it could potentially help high dimensional problems like inference in graphical models. We show a general framework for unsupervised learning based on optimal transport and sparse regularization. Optimal transport presents interesting challenges from an optimization point of view with its simplex constraints on the rows and columns of the transport plan. We show one way to formulate efficient optimization problems inspired by optimal transport. This could be done by imposing only one set of the simplex constraints and by imposing structure on the transport plan through sparse regularization. We show how unsupervised learning algorithms like exemplar clustering, center based clustering and kernel PCA could fit into this framework based on different forms of regularization. We especially demonstrate a promising approach to address the pre-image problem in kernel PCA. Several methods have been proposed over the years, which generally assume certain types of kernels or have too many hyper-parameters or make restrictive approximations of the underlying geometry. We present a more general method, with only one hyper-parameter to tune and with some interesting geometric properties. From an optimization point of view, we show how to compute the gradient of a smooth version of the Schatten p-norm and how it can be used within a majorization-minimization scheme. Finally, we present results from our various experiments
Pascal, Barbara. "Estimation régularisée d'attributs fractals par minimisation convexe pour la segmentation de textures : formulations variationnelles conjointes, algorithmes proximaux rapides et sélection non supervisée des paramètres de régularisation; Applications à l'étude du frottement solide et de la microfluidique des écoulements multiphasiques." Thesis, Lyon, 2020. http://www.theses.fr/2020LYSEN042.
In this doctoral thesis several scale-free texture segmentation procedures based on two fractal attributes, the Hölder exponent, measuring the local regularity of a texture, and local variance, are proposed.A piecewise homogeneous fractal texture model is built, along with a synthesis procedure, providing images composed of the aggregation of fractal texture patches with known attributes and segmentation. This synthesis procedure is used to evaluate the proposed methods performance.A first method, based on the Total Variation regularization of a noisy estimate of local regularity, is illustrated and refined thanks to a post-processing step consisting in an iterative thresholding and resulting in a segmentation.After evidencing the limitations of this first approach, deux segmentation methods, with either "free" or "co-located" contours, are built, taking in account jointly the local regularity and the local variance.These two procedures are formulated as convex nonsmooth functional minimization problems.We show that the two functionals, with "free" and "co-located" penalizations, are both strongly-convex. and compute their respective strong convexity moduli.Several minimization schemes are derived, and their convergence speed are compared.The segmentation performance of the different methods are evaluated over a large amount of synthetic data in configurations of increasing difficulty, as well as on real world images, and compared to state-of-the-art procedures, including convolutional neural networks.An application for the segmentation of multiphasic flow through a porous medium experiment images is presented.Finally, a strategy for automated selection of the hyperparameters of the "free" and "co-located" functionals is built, inspired from the SURE estimator of the quadratic risk
Velcin, Julien. "Extraction automatique de stéréotypes à partir de données symboliques et lacunaires." Paris 6, 2005. http://www.theses.fr/2005PA066465.
Wynen, Daan. "Une représentation archétypale de style artistique : résumer et manipuler des stylesartistiques d'une façon interprétable." Thesis, Université Grenoble Alpes, 2020. http://www.theses.fr/2020GRALM066.
In this thesis we study the representations used to describe and manipulate artistic style of visual arts.In the neural style transfer literature and related strains of research, different representations have been proposed, but in recent years the by far dominant representations of artistic style in the computer vision community have been those learned by deep neural networks, trained on natural images.We build on these representations with the dual goal of summarizing the artistic styles present in large collections of digitized artworks, as well as manipulating the styles of images both natural and artistic.To this end, we propose a concise and intuitive representation based on archetypal analysis, a classic unsupervised learning method with properties that make it especially suitable for the task. We demonstrate how this archetypal representation of style can be used to discover and describe, in an interpretable way, which styles are present in a large collection.This enables the exploration of styles present in a collection from different angles; different ways of visualizing the information allow for different questions to be asked.These can be about a style that was identified across artworks, about the style of a particular artwork, or more broadly about how the styles that were identified relate to one another.We apply our analysis to a collection of artworks obtained from WikiArt, an online collection effort of visual arts driven by volunteers. This dataset also includes metadata such as artist identies, genre, and style of the artworks. We use this metadata for further analysis of the archetypal style representation along biographic lines of artists and with an eye on the relationships within groups of artists
Chafik, Sanaa. "Machine learning techniques for content-based information retrieval." Thesis, Université Paris-Saclay (ComUE), 2017. http://www.theses.fr/2017SACLL008/document.
The amount of media data is growing at high speed with the fast growth of Internet and media resources. Performing an efficient similarity (nearest neighbor) search in such a large collection of data is a very challenging problem that the scientific community has been attempting to tackle. One of the most promising solutions to this fundamental problem is Content-Based Media Retrieval (CBMR) systems. The latter are search systems that perform the retrieval task in large media databases based on the content of the data. CBMR systems consist essentially of three major units, a Data Representation unit for feature representation learning, a Multidimensional Indexing unit for structuring the resulting feature space, and a Nearest Neighbor Search unit to perform efficient search. Media data (i.e. image, text, audio, video, etc.) can be represented by meaningful numeric information (i.e. multidimensional vector), called Feature Description, describing the overall content of the input data. The task of the second unit is to structure the resulting feature descriptor space into an index structure, where the third unit, effective nearest neighbor search, is performed.In this work, we address the problem of nearest neighbor search by proposing three Content-Based Media Retrieval approaches. Our three approaches are unsupervised, and thus can adapt to both labeled and unlabeled real-world datasets. They are based on a hashing indexing scheme to perform effective high dimensional nearest neighbor search. Unlike most recent existing hashing approaches, which favor indexing in Hamming space, our proposed methods provide index structures adapted to a real-space mapping. Although Hamming-based hashing methods achieve good accuracy-speed tradeoff, their accuracy drops owing to information loss during the binarization process. By contrast, real-space hashing approaches provide a more accurate approximation in the mapped real-space as they avoid the hard binary approximations.Our proposed approaches can be classified into shallow and deep approaches. In the former category, we propose two shallow hashing-based approaches namely, "Symmetries of the Cube Locality Sensitive Hashing" (SC-LSH) and "Cluster-based Data Oriented Hashing" (CDOH), based respectively on randomized-hashing and shallow learning-to-hash schemes. The SC-LSH method provides a solution to the space storage problem faced by most randomized-based hashing approaches. It consists of a semi-random scheme reducing partially the randomness effect of randomized hashing approaches, and thus the memory storage problem, while maintaining their efficiency in structuring heterogeneous spaces. The CDOH approach proposes to eliminate the randomness effect by combining machine learning techniques with the hashing concept. The CDOH outperforms the randomized hashing approaches in terms of computation time, memory space and search accuracy.The third approach is a deep learning-based hashing scheme, named "Unsupervised Deep Neuron-per-Neuron Hashing" (UDN2H). The UDN2H approach proposes to index individually the output of each neuron of the top layer of a deep unsupervised model, namely a Deep Autoencoder, with the aim of capturing the high level individual structure of each neuron output.Our three approaches, SC-LSH, CDOH and UDN2H, were proposed sequentially as the thesis was progressing, with an increasing level of complexity in terms of the developed models, and in terms of the effectiveness and the performances obtained on large real-world datasets
Besson, Lilian. "Multi-Players Bandit Algorithms for Internet of Things Networks." Thesis, CentraleSupélec, 2019. http://www.theses.fr/2019CSUP0005.
In this PhD thesis, we study wireless networks and reconfigurable end-devices that can access Cognitive Radio networks, in unlicensed bands and without central control. We focus on Internet of Things networks (IoT), with the objective of extending the devices’ battery life, by equipping them with low-cost but efficient machine learning algorithms, in order to let them automatically improve the efficiency of their wireless communications. We propose different models of IoT networks, and we show empirically on both numerical simulations and real-world validation the possible gain of our methods, that use Reinforcement Learning. The different network access problems are modeled as Multi-Armed Bandits (MAB), but we found that analyzing the realistic models was intractable, because proving the convergence of many IoT devices playing a collaborative game, without communication nor coordination is hard, when they all follow random activation patterns. The rest of this manuscript thus studies two restricted models, first multi-players bandits in stationary problems, then non-stationary single-player bandits. We also detail another contribution, SMPyBandits, our open-source Python library for numerical MAB simulations, that covers all the studied models and more
Flandin, Guillaume. "Utilisation d'informations géométriques pour l'analyse statistique des données d'IRM fonctionnelle." Phd thesis, Université de Nice Sophia-Antipolis, 2004. http://tel.archives-ouvertes.fr/tel-00633520.
Chen, Yuxin. "Apprentissage interactif de mots et d'objets pour un robot humanoïde." Thesis, Université Paris-Saclay (ComUE), 2017. http://www.theses.fr/2017SACLY003/document.
Future applications of robotics, especially personal service robots, will require continuous adaptability to the environment, and particularly the ability to recognize new objects and learn new words through interaction with humans. Though having made tremendous progress by using machine learning, current computational models for object detection and representation still rely heavily on good training data and ideal learning supervision. In contrast, two year old children have an impressive ability to learn to recognize new objects and at the same time to learn the object names during interaction with adults and without precise supervision. Therefore, following the developmental robotics approach, we develop in the thesis learning approaches for objects, associating their names and corresponding features, inspired by the infants' capabilities, in particular, the ambiguous interaction with humans, inspired by the interaction that occurs between children and parents.The general idea is to use cross-situational learning (finding the common points between different presentations of an object or a feature) and to implement multi-modal concept discovery based on two latent topic discovery approaches : Non Negative Matrix Factorization (NMF) and Latent Dirichlet Association (LDA). Based on vision descriptors and sound/voice inputs, the proposed approaches will find the underlying regularities in the raw dataflow to produce sets of words and their associated visual meanings (eg. the name of an object and its shape, or a color adjective and its correspondence in images). We developed a complete approach based on these algorithms and compared their behavior in front of two sources of uncertainties: referential ambiguities, in situations where multiple words are given that describe multiple objects features; and linguistic ambiguities, in situations where keywords we intend to learn are merged in complete sentences. This thesis highlights the algorithmic solutions required to be able to perform efficient learning of these word-referent associations from data acquired in a simplified but realistic acquisition setup that made it possible to perform extensive simulations and preliminary experiments in real human-robot interactions. We also gave solutions for the automatic estimation of the number of topics for both NMF and LDA.We finally proposed two active learning strategies, Maximum Reconstruction Error Based Selection (MRES) and Confidence Based Exploration (CBE), to improve the quality and speed of incremental learning by letting the algorithms choose the next learning samples. We compared the behaviors produced by these algorithms and show their common points and differences with those of humans in similar learning situations
Ho, Dinh Khanh. "Gestion des ressources et de l’énergie orientée qualité de service pour les systèmes robotiques mobiles autonomes." Thesis, Université Côte d'Azur, 2020. http://www.theses.fr/2020COAZ4000.
Mobile robotic systems are becoming more and more complex with the integration of advanced sensing and acting components and functionalities to perform the real required missions. For these technical systems, the requirements are divided into two categories: functional and non-functional requirements. While functional requirements represent what the robot must do to accomplish the mission, non-functional requirements represent how the robot performs the mission. Thus, the quality of service and energy efficiency of a robotic mission are classified in this category. The autonomy of these systems is fully achieved when both functional and non-functional requirements are guaranteed without any human intervention or any external control. However, these mobile systems are naturally confronted with resource availability and energy capacity constraints, particularly in the context of long-term missions, these constraints become more critical. In addition, the performance of these systems is also influenced by unexpected and unstructured environmental conditions in which they interact. The management of resources and energy during operation is therefore a challenge for autonomous mobile robots in order to guarantee the desired performance objectives while respecting constraints. In this context, the ability of the robotic system to become aware of its own internal behaviors and physical environment and to adapt to these dynamic circumstances becomes important.This thesis focuses on the quality of service and energy efficiency of mobile robotic systems and proposes a hierarchical run-time management in order to guarantee these non-functional objectives of each robotic mission. At the local management level of each robotic mission, a Mission Manager employs a reinforcement learning-based decision-making mechanism to automatically reconfigure certain key mission-specific parameters to minimize the level of violation of required performance and energy objectives. At the global management level of the whole system, a Multi-Mission Manager leveraged rule-based decision-making and case-based reasoning techniques monitors the system's resources and the responses of Mission Managers in order to decide to reallocate the energy budget, regulate the quality of service and trigger the online learning for each robotic mission.The proposed methodology has been successfully prototyped and validated in a simulation environment and the run-time management framework is also integrated into our real mobile robotic system based on a Pioneer-3DX mobile base equipped with an embedded NVIDIA Jetson Xavier platform
Besson, Florent. "Integrating PET-MR data for a multiparametric approach of tumour heterogeneity in non-small-cell lung cancer (NSCLC)." Thesis, université Paris-Saclay, 2020. http://www.theses.fr/2020UPASS081.
Tumor heterogeneity is an important factor of progression and resistance to treatment. Multiparametric PET-MRI imaging offers unique opportunities to characterize biological cellular processes, but has never been evaluated at the regional level in Non-Small Cell Lung Cancer (NSCLC), the leading cause of oncological death. A simultaneous dynamic multiparametric 18F-FDG PET-MRI approach has been developed to this end. This approach required the “in-house” implementation of the reference absolute PET quantitative method of glucose metabolism (Sokoloff's tri-compartmental model); the development of a method for correcting geometric distortions in diffusion weighted imaging, validated on phantom and clinically tested; the phantom validation of quantitative MRI methods (T1/T2 relaxometry), also clinically tested; and the "in-house" implementation of the Tofts compartmental model (extended version) for the evaluation of tumor vascularization by dynamic perfusion MRI. The results of our work, performed at the regional intra-tumor level, illustrate the heterogeneity of the regional interlinks between glucose metabolism and vascularization in NSCLC, two fundamental biological hallmarks of tumor progression, and show that an unsupervised tumor partitioning by Gaussian mixture model, integrating all the PET-MRI biomarkers of this project, individualizes 3 types of supervoxels, whose biological signature can be predicted with 97% accuracy by 4 dominant PET-MRI biomarkers, revealed by metaheuristic machine learning methods
Laquitaine, Steeve. "Les bases neuronales de l’apprentissage décisionnel au sein des ganglions de la base : étude électrophysiologique et comportementale chez le primate non humain." Thesis, Bordeaux 2, 2010. http://www.theses.fr/2010BOR21743/document.
A fundamental question in neuroscience, as well as in various fields such as economics, psychology and sociology, concerns the decision making processes by which animals and humans select actions based on reward and punishment. Both decision making processes and their neural basis are still poorly understood. Also, both human and animals often make suboptimal decisions in many tasks studied. Our first aim is to improve the understanding of why such sub-optimal decisions are made. Also, the alteration of decision making processes causes diseases, the understanding of whose mechanisms is essential in developing better treatment strategies. In this report, we propose a new approach which consists in extracting the neural substrates of choice behavior heterogeneity in between sessions. Our results show that although primates learn on average to identify the best option and converge to an optimal policy in a consequent number of sessions, they fail on average to optimize their behavior. We revealed that this suboptimal behavior was characterized by an unexpected high behavioral heterogeneity during the task that was due to the creation of irrelevant preferences by the monkeys. We finally show that although a few neurons of the putamen encode the action value, their contribution to the overall population activity is weak. Putamen activity rather reflects the futures performances and predicts the creation of rational and irrational behaviors
Cruz, Cavalcanti Yanna. "Factor analysis of dynamic PET images." Thesis, Toulouse, INPT, 2018. http://www.theses.fr/2018INPT0078/document.
Thanks to its ability to evaluate metabolic functions in tissues from the temporal evolution of a previously injected radiotracer, dynamic positron emission tomography (PET) has become an ubiquitous analysis tool to quantify biological processes. Several quantification techniques from the PET imaging literature require a previous estimation of global time-activity curves (TACs) (herein called \textit{factors}) representing the concentration of tracer in a reference tissue or blood over time. To this end, factor analysis has often appeared as an unsupervised learning solution for the extraction of factors and their respective fractions in each voxel. Inspired by the hyperspectral unmixing literature, this manuscript addresses two main drawbacks of general factor analysis techniques applied to dynamic PET. The first one is the assumption that the elementary response of each tissue to tracer distribution is spatially homogeneous. Even though this homogeneity assumption has proven its effectiveness in several factor analysis studies, it may not always provide a sufficient description of the underlying data, in particular when abnormalities are present. To tackle this limitation, the models herein proposed introduce an additional degree of freedom to the factors related to specific binding. To this end, a spatially-variant perturbation affects a nominal and common TAC representative of the high-uptake tissue. This variation is spatially indexed and constrained with a dictionary that is either previously learned or explicitly modelled with convolutional nonlinearities affecting non-specific binding tissues. The second drawback is related to the noise distribution in PET images. Even though the positron decay process can be described by a Poisson distribution, the actual noise in reconstructed PET images is not expected to be simply described by Poisson or Gaussian distributions. Therefore, we propose to consider a popular and quite general loss function, called the $\beta$-divergence, that is able to generalize conventional loss functions such as the least-square distance, Kullback-Leibler and Itakura-Saito divergences, respectively corresponding to Gaussian, Poisson and Gamma distributions. This loss function is applied to three factor analysis models in order to evaluate its impact on dynamic PET images with different reconstruction characteristics
Dekhtiar, Jonathan. "Deep Learning and unsupervised learning to automate visual inspection in the manufacturing industry." Thesis, Compiègne, 2019. http://www.theses.fr/2019COMP2513.
Although studied since 1970, automatic visual inspection on production lines still struggles to be applied on a large scale and at low cost. The methods used depend greatly on the availability of domain experts. This inevitably leads to increased costs and reduced flexibility in the methods used. Since 2012, advances in the field of Deep Learning have enabled many advances in this direction, particularly thanks to convolutional neura networks that have achieved near-human performance in many areas associated with visual perception (e.g. object recognition and detection, etc.). This thesis proposes an unsupervised approach to meet the needs of automatic visual inspection. This method, called AnoAEGAN, combines adversarial learning and the estimation of a probability density function. These two complementary approaches make it possible to jointly estimate the pixel-by-pixel probability of a visual defect on an image. The model is trained from a very limited number of images (i.e. less than 1000 images) without using expert knowledge to "label" the data beforehand. This method allows increased flexibility with a limited training time and therefore great versatility, demonstrated on ten different tasks without any modification of the model. This method should reduce development costs and the time required to deploy in production. This method can also be deployed in a complementary way to a supervised approach in order to benefit from the advantages of each approach
Gulikers, Lennart. "Sur deux problèmes d’apprentissage automatique : la détection de communautés et l’appariement adaptatif." Thesis, Paris Sciences et Lettres (ComUE), 2017. http://www.theses.fr/2017PSLEE062/document.
In this thesis, we study two problems of machine learning: (I) community detection and (II) adaptive matching. I) It is well-known that many networks exhibit a community structure. Finding those communities helps us understand and exploit general networks. In this thesis we focus on community detection using so-called spectral methods based on the eigenvectors of carefully chosen matrices. We analyse their performance on artificially generated benchmark graphs. Instead of the classical Stochastic Block Model (which does not allow for much degree-heterogeneity), we consider a Degree-Corrected Stochastic Block Model (DC-SBM) with weighted vertices, that is able to generate a wide class of degree sequences. We consider this model in both a dense and sparse regime. In the dense regime, we show that an algorithm based on a suitably normalized adjacency matrix correctly classifies all but a vanishing fraction of the nodes. In the sparse regime, we show that the availability of only a small amount of information entails the existence of an information-theoretic threshold below which no algorithm performs better than random guess. On the positive side, we show that an algorithm based on the non-backtracking matrix works all the way down to the detectability threshold in the sparse regime, showing the robustness of the algorithm. This follows after a precise characterization of the non-backtracking spectrum of sparse DC-SBM's. We further perform tests on well-known real networks. II) Online two-sided matching markets such as Q&A forums and online labour platforms critically rely on the ability to propose adequate matches based on imperfect knowledge of the two parties to be matched. We develop a model of a task / server matching system for (efficient) platform operation in the presence of such uncertainty. For this model, we give a necessary and sufficient condition for an incoming stream of tasks to be manageable by the system. We further identify a so-called back-pressure policy under which the throughput that the system can handle is optimized. We show that this policy achieves strictly larger throughput than a natural greedy policy. Finally, we validate our model and confirm our theoretical findings with experiments based on user-contributed content on an online platform
Vanhoutte, Matthieu. "Caractérisation par imagerie TEP 18F-FDG de la maladie d’Alzheimer à début précoce." Thesis, Lille 2, 2018. http://www.theses.fr/2018LIL2S026/document.
Alzheimer’s disease (AD) is the most common form of neurodegenerative dementia, characterized at 95% by late-onset forms (LOAD) which present episodic memory impairments and progress slowly. However, 5% of AD patients have an early-onset form (EOAD) of the disease whose onset begins before 65. Although the lesion substratum is similar between EOAD and LOAD, EOAD has more severe neuritic plaque deposits, neurofibrillary tangles and brain atrophy. Moreover, EOAD is more heterogeneous than LOAD, because even if most of the impairments are about episodic memory there is a high proportion of atypical forms impaired in language, visuospatial or executive functions. Although many 18F-FDG PET studies allowed to metabolically characterize EOAD compared to LOAD or healthy controls group, very few differentiated typical from atypical forms. In this thesis, we examined 18F-FDG PET data, complemented by structural MRI, in order to improve characterization and comprehension of typical and atypical forms of EOAD. Following a first harmonization work between 18F-FDG PET reconstructions from both GE and Siemens scanners used for the acquisition of patient data, our second aim was to study at baseline on the whole brain hypometabolic patterns characterizing the clinical forms of EOAD and their correlations with neuropsychological performance. This work showed that each clinical form of EOAD was characterized by specific hypometabolic patterns highly correlated with clinical symptoms and neuropsychological performance of the associated cognitive domain. Then, we focused on the 3-year hypometabolism progression on the cortical surface according typical or atypical forms of EOAD. Although similar patterns of hypometabolism evolution between typical and atypical forms were observed in parietal cortices, atypical only showed a more severe reduction of metabolism in lateral orbitofrontal cortices associated with more severe cognitive declines. Temporally, the results suggest that hypometabolism in typical forms would progress according to an anterior-to-posterior axis coherently with Braak and Braak stages, whereas in atypical forms hypometabolism would progress according a posterior-to-anterior axis. Taken together, results consolidate the hypothesis of a different tau distribution in terms of burden and temporal evolution between both forms of EOAD. Our last goal was to determine the discriminative power of 18F-FDG PET data, alone or combined to structural MRI data, in order to automatically classify in a supervised manner EOAD patients into typical or atypical form. We applied machine learning algorithms combined to cross-validation methods to assess influence of some components on classification performances. Maximum balanced accuracies equal to 80.8% in monomodal 18F-FDG PET and 92.4% in multimodal 18F-FDG PET/T1 MRI were obtained, validating 18F-FDG PET as a sensible biomarker of EOAD and highlighting the incontestable contribution of multimodality. In conclusion, our works allowed a better characterization and comprehension of clinical forms of EOAD, paving the way to personalized patient management and more effective treatments for these distinct clinical forms
De, lozzo Matthias. "Modèles de substitution spatio-temporels et multifidélité : Application à l'ingénierie thermique." Thesis, Toulouse, INSA, 2013. http://www.theses.fr/2013ISAT0027/document.
This PhD thesis deals with the construction of surrogate models in transient and steady states in the context of thermal simulation, with a few observations and many outputs.First, we design a robust construction of recurrent multilayer perceptron so as to approach a spatio-temporal dynamic. We use an average of neural networks resulting from a cross-validation procedure, whose associated data splitting allows to adjust the parameters of these models thanks to a test set without any information loss. Moreover, the construction of this perceptron can be distributed according to its outputs. This construction is applied to the modelling of the temporal evolution of the temperature at different points of an aeronautical equipment.Then, we proposed a mixture of Gaussian process models in a multifidelity framework where we have a high-fidelity observation model completed by many observation models with lower and no comparable fidelities. A particular attention is paid to the specifications of trends and adjustement coefficients present in these models. Different kriging and co-krigings models are put together according to a partition or a weighted aggregation based on a robustness measure associated to the most reliable design points. This approach is used in order to model the temperature at different points of the equipment in steady state.Finally, we propose a penalized criterion for the problem of heteroscedastic regression. This tool is build in the case of projection estimators and applied with the Haar wavelet. We also give some numerical results for different noise specifications and possible dependencies in the observations
Racah, Evan. "Unsupervised representation learning in interactive environments." Thèse, 2019. http://hdl.handle.net/1866/23788.
Extracting a representation of all the high-level factors of an agent’s state from level-level sensory information is an important, but challenging task in machine learning. In this thesis, we will explore several unsupervised approaches for learning these state representations. We apply and analyze existing unsupervised representation learning methods in reinforcement learning environments, as well as contribute our own evaluation benchmark and our own novel state representation learning method. In the first chapter, we will overview and motivate unsupervised representation learning for machine learning in general and for reinforcement learning. We will then introduce a relatively new subfield of representation learning: self-supervised learning. We will then cover two core representation learning approaches, generative methods and discriminative methods. Specifically, we will focus on a collection of discriminative representation learning methods called contrastive unsupervised representation learning (CURL) methods. We will close the first chapter by detailing various approaches for evaluating the usefulness of representations. In the second chapter, we will present a workshop paper, where we evaluate a handful of off-the-shelf self-supervised methods in reinforcement learning problems. We discover that the performance of these representations depends heavily on the dynamics and visual structure of the environment. As such, we determine that a more systematic study of environments and methods is required. Our third chapter covers our second article, Unsupervised State Representation Learning in Atari, where we try to execute a more thorough study of representation learning methods in RL as motivated by the second chapter. To facilitate a more thorough evaluation of representations in RL we introduce a benchmark of 22 fully labelled Atari games. In addition, we choose the representation learning methods for comparison in a more systematic way by focusing on comparing generative methods with contrastive methods, instead of the less systematically chosen off-the-shelf methods from the second chapter. Finally, we introduce a new contrastive method, ST-DIM, which excels at the 22 Atari games.
Payette, François. "Apprentissage semi-supervisé par réduction de dimensionnalité non linéaire." Thèse, 2004. http://hdl.handle.net/1866/16641.
Lavallée, Jean-François. "Moranapho : apprentissage non supervisé de la morphologie d'une langue par généralisation de relations analogiques." Thèse, 2010. http://hdl.handle.net/1866/4524.
Recently, we have witnessed a growing interest in applying the concept of formal analogy to unsupervised morphology acquisition. The attractiveness of this concept lies in its parallels with the mental process involved in the creation of new words based on morphological relations existing in the language. However, the use of formal analogy remain marginal partly due to their high computational cost. In this document, we present Moranapho, a graph-based system founded on the concept of formal analogy. Our participation in the 2009 Morpho Challenge (Kurimo:10) and our subsequent experiments demonstrate that the performance of Moranapho are favorably comparable to the state-of-the-art. We studied the influence of some of its components on the quality of the morphological analysis produced as well. Finally, we will discuss our findings based on well-established theories in the field of linguistics. This allows us to provide some predictions on the successes and failures of our system when applied to languages other than those tested in our experiments.
Lajoie, Isabelle. "Apprentissage de représentations sur-complètes par entraînement d’auto-encodeurs." Thèse, 2009. http://hdl.handle.net/1866/3768.
Progress in the machine learning domain allows computational system to address more and more complex tasks associated with vision, audio signal or natural language processing. Among the existing models, we find the Artificial Neural Network (ANN), whose popularity increased suddenly with the recent breakthrough of Hinton et al. [22], that consists in using Restricted Boltzmann Machines (RBM) for performing an unsupervised, layer by layer, pre-training initialization, of a Deep Belief Network (DBN), which enables the subsequent successful supervised training of such architecture. Since this discovery, researchers studied the efficiency of other similar pre-training strategies such as the stacking of traditional auto-encoder (SAE) [5, 38] and the stacking of denoising auto-encoder (SDAE) [44]. This is the context in which the present study started. After a brief introduction of the basic machine learning principles and of the pre-training methods used until now with RBM, AE and DAE modules, we performed a series of experiments to deepen our understanding of pre-training with SDAE, explored its different proprieties and explored variations on the DAE algorithm as alternative strategies to initialize deep networks. We evaluated the sensitivity to the noise level, and influence of number of layers and number of hidden units on the generalization error obtained with SDAE. We experimented with other noise types and saw improved performance on the supervised task with the use of pepper and salt noise (PS) or gaussian noise (GS), noise types that are more justified then the one used until now which is masking noise (MN). Moreover, modifying the algorithm by imposing an emphasis on the corrupted components reconstruction during the unsupervised training of each different DAE showed encouraging performance improvements. Our work also allowed to reveal that DAE was capable of learning, on naturals images, filters similar to those found in V1 cells of the visual cortex, that are in essence edges detectors. In addition, we were able to verify that the learned representations of SDAE, are very good characteristics to be fed to a linear or gaussian support vector machine (SVM), considerably enhancing its generalization performance. Also, we observed that, alike DBN, and unlike SAE, the SDAE had the potential to be used as a good generative model. As well, we opened the door to novel pre-training strategies and discovered the potential of one of them : the stacking of renoising auto-encoders (SRAE).
Paquette, Philip. "No Press Diplomacy." Thèse, 2019. http://hdl.handle.net/1866/22535.
Schwarzer, Max. "Data-efficient reinforcement learning with self-predictive representations." Thesis, 2020. http://hdl.handle.net/1866/25105.
Data efficiency remains a key challenge in deep reinforcement learning. Although modern techniques have been shown to be capable of attaining high performance in extremely complex tasks, including strategy games such as StarCraft, Chess, Shogi, and Go as well as in challenging visual domains such as Atari games, doing so generally requires enormous amounts of interactional data, limiting how broadly reinforcement learning can be applied. In this thesis, we propose SPR, a method drawing from recent advances in self-supervised representation learning designed to enhance the data efficiency of deep reinforcement learning agents. We evaluate this method on the Atari Learning Environment, and show that it dramatically improves performance with limited computational overhead. When given roughly the same amount of learning time as human testers, a reinforcement learning agent augmented with SPR achieves super-human performance on 7 out of 26 games, an increase of 350% over the previous state of the art, while also strongly improving mean and median performance. We also evaluate this method on a set of continuous control tasks, showing substantial improvements over previous methods. Chapter 1 introduces concepts necessary to understand the work presented, including overviews of Deep Reinforcement Learning and Self-Supervised Representation learning. Chapter 2 contains a detailed description of our contributions towards leveraging self-supervised representation learning to improve data-efficiency in reinforcement learning. Chapter 3 provides some conclusions drawn from this work, including a number of proposals for future work.
Léonard, Nicholas. "Distributed conditional computation." Thèse, 2014. http://hdl.handle.net/1866/11954.
The objective of this paper is to present different applications of the distributed conditional computation research program. It is hoped that these applications and the theory presented here will lead to a general solution of the problem of artificial intelligence, especially with regard to the need for efficiency. The vision of distributed conditional computation is to accelerate the evaluation and training of deep models which is very different from the usual objective of improving its generalization and optimization capacity. The work presented here has close ties with mixture of experts models. In Chapter 2, we present a new deep learning algorithm that uses a form of reinforcement learning on a novel neural network decision tree model. We demonstrate the need for a balancing constraint to keep the distribution of examples to experts uniform and to prevent monopolies. To make the calculation efficient, the training and evaluation are constrained to be sparse by using a gater that samples experts from a multinomial distribution given examples. In Chapter 3 we present a new deep model consisting of a sparse representation divided into segments of experts. A neural network language model is constructed from blocks of sparse transformations between these expert segments. The block-sparse operation is implemented for use on graphics cards. Its speed is compared with two dense operations of the same caliber to demonstrate and measure the actual efficiency gain that can be obtained. A deep model using these block-sparse operations controlled by a distinct gater is trained on a dataset of one billion words. A new algorithm for data partitioning (clustering) is applied to a set of words to organize the output layer of a language model into a conditional hierarchy, thereby making it much more efficient. The work presented in this thesis is central to the vision of distributed conditional computation as issued by Yoshua Bengio. It attempts to apply research in the area of mixture of experts to deep models to improve their speed and their optimization capacity. We believe that the theory and experiments of this thesis are an important step on the path to distributed conditional computation because it provides a good framework for the problem, especially concerning competitiveness inherent to systems of experts.
Parent-Lévesque, Jérôme. "Towards deep unsupervised inverse graphics." Thesis, 2020. http://hdl.handle.net/1866/25467.
A long standing goal of computer vision is to infer the underlying 3D content in a scene from a single photograph, a task known as inverse graphics. Machine learning has, in recent years, enabled many approaches to make great progress towards solving this problem. However, most approaches rely on 3D supervision data which is expensive and sometimes impossible to obtain and therefore limits the learning capabilities of such work. In this work, we explore the deep unsupervised inverse graphics training pipeline and propose two methods based on distinct 3D representations and associated differentiable rendering algorithms: namely surfels and a novel Voronoi-based representation. In the first method based on surfels, we show that, while effective at maintaining view-consistency, producing view-dependent surfels using a learned depth map results in ambiguities as the mapping between depth map and rendering is non-bijective. In our second method, we introduce a novel 3D representation based on Voronoi diagrams which models objects/scenes both explicitly and implicitly simultaneously, thereby combining the benefits of both. We show how this representation can be used in both a supervised and unsupervised context and discuss its advantages compared to traditional 3D representations.
Hosseini, Seyedarian. "Towards learning sentence representation with self-supervision." Thèse, 2019. http://hdl.handle.net/1866/23784.
In chapter 1, we introduce the basics of feed forward neural networks and recurrent neural networks. The chapter continues with the discussion of the backpropagation algorithm to train feed forward neural networks, and the backpropagation through time algorithm to train recurrent neural networks. We also discuss three different approaches in learning representations, namely supervised learning, unsupervised learning, and a relatively new approach called self-supervised learning. In chapter 2, we talk about the fundamentals of deep natural language processing. Specifically, we cover word representations, sentence representations, and language modelling. We focus on the evaluation and current state of the literature for these concepts. We close the chapter by discussing large scale pre-training and transfer learning in language. In chapter 3, we investigate a set of self-supervised tasks that take advantage of noise contrastive estimation in order to learn sentence representations using unlabeled data. We train our model on a large corpora and evaluate our learned sentence representations on a set of downstream natural language tasks from the SentEval framework. Our model trained on the proposed tasks outperforms unsupervised methods on a subset of tasks from SentEval. In chapter 4, we introduce a memory augmented model called Ordered Memory with several improvements over traditional stack-augmented recurrent neural networks. We introduce a new Stick-breaking attention mechanism inspired by Ordered Neurons [Shen et.al., 2019] to write in and erase from the memory. A new Gated Recursive Cell is also introduced to compose low level representations into higher level ones. We show that this model performs well on the logical inference task and the ListOps task, and it also shows strong generalization properties in these tasks. Finally, we evaluate our model on the SST (Stanford Sentiment Treebank) tasks (binary and fine-grained) and report results that are comparable with state-of-the-art on these tasks.
Goyette, Kyle. "On two sequential problems : the load planning and sequencing problem and the non-normal recurrent neural network." Thesis, 2020. http://hdl.handle.net/1866/24314.
Le travail de cette thèse est divisé en deux parties. La première partie traite du problème de planification et de séquencement des chargements de conteneurs sur des wagons, un problème opérationnel rencontré dans de nombreux terminaux ferroviaires intermodaux. Dans ce problème, les conteneurs doivent être affectés à une plate-forme sur laquelle un ou deux conteneurs seront chargés et l'ordre de chargement doit être déterminé. Ces décisions sont prises dans le but de minimiser les coûts associés à la manutention des conteneurs, ainsi que de minimiser le coût des conteneurs non chargés. La version déterministe du problème peut être formulé comme un problème de plus court chemin sur un graphe ordonné. Ce problème est difficile à résoudre en raison de la grande taille du graphe. Nous proposons une heuristique en deux étapes basée sur l'algorithme Iterative Deepening A* pour calculer des solutions au problème de planification et de séquencement de la charge dans un budget de cinq minutes. Ensuite, nous illustrons également comment un algorithme d'apprentissage Deep Q peut être utilisé pour résoudre heuristiquement le même problème. La deuxième partie de cette thèse examine les modèles séquentiels en apprentissage profond. Une stratégie récente pour contourner le problème de gradient qui explose et disparaît dans les réseaux de neurones récurrents (RNN) consiste à imposer des matrices de poids récurrentes orthogonales ou unitaires. Bien que cela assure une dynamique stable pendant l'entraînement, cela se fait au prix d'une expressivité réduite en raison de la variété limitée des transformations orthogonales. Nous proposons une paramétrisation des RNN, basée sur la décomposition de Schur, qui atténue les problèmes de gradient, tout en permettant des matrices de poids récurrentes non orthogonales dans le modèle.
Dumoulin, Vincent. "Representation Learning for Visual Data." Thèse, 2018. http://hdl.handle.net/1866/21140.
Desjardins, Guillaume. "Improving sampling, optimization and feature extraction in Boltzmann machines." Thèse, 2013. http://hdl.handle.net/1866/10550.
Despite the current widescale success of deep learning in training large scale hierarchical models through supervised learning, unsupervised learning promises to play a crucial role towards solving general Artificial Intelligence, where agents are expected to learn with little to no supervision. The work presented in this thesis tackles the problem of unsupervised feature learning and density estimation, using a model family at the heart of the deep learning phenomenon: the Boltzmann Machine (BM). We present contributions in the areas of sampling, partition function estimation, optimization and the more general topic of invariant feature learning. With regards to sampling, we present a novel adaptive parallel tempering method which dynamically adjusts the temperatures under simulation to maintain good mixing in the presence of complex multi-modal distributions. When used in the context of stochastic maximum likelihood (SML) training, the improved ergodicity of our sampler translates to increased robustness to learning rates and faster per epoch convergence. Though our application is limited to BM, our method is general and is applicable to sampling from arbitrary probabilistic models using Markov Chain Monte Carlo (MCMC) techniques. While SML gradients can be estimated via sampling, computing data likelihoods requires an estimate of the partition function. Contrary to previous approaches which consider the model as a black box, we provide an efficient algorithm which instead tracks the change in the log partition function incurred by successive parameter updates. Our algorithm frames this estimation problem as one of filtering performed over a 2D lattice, with one dimension representing time and the other temperature. On the topic of optimization, our thesis presents a novel algorithm for applying the natural gradient to large scale Boltzmann Machines. Up until now, its application had been constrained by the computational and memory requirements of computing the Fisher Information Matrix (FIM), which is square in the number of parameters. The Metric-Free Natural Gradient algorithm (MFNG) avoids computing the FIM altogether by combining a linear solver with an efficient matrix-vector operation. The method shows promise in that the resulting updates yield faster per-epoch convergence, despite being slower in terms of wall clock time. Finally, we explore how invariant features can be learnt through modifications to the BM energy function. We study the problem in the context of the spike & slab Restricted Boltzmann Machine (ssRBM), which we extend to handle both binary and sparse input distributions. By associating each spike with several slab variables, latent variables can be made invariant to a rich, high dimensional subspace resulting in increased invariance in the learnt representation. When using the expected model posterior as input to a classifier, increased invariance translates to improved classification accuracy in the low-label data regime. We conclude by showing a connection between invariance and the more powerful concept of disentangling factors of variation. While invariance can be achieved by pooling over subspaces, disentangling can be achieved by learning multiple complementary views of the same subspace. In particular, we show how this can be achieved using third-order BMs featuring multiplicative interactions between pairs of random variables.