Tesis sobre el tema "Apprentissage de règles à des PLM"
Crea una cita precisa en los estilos APA, MLA, Chicago, Harvard y otros
Consulte los 50 mejores tesis para su investigación sobre el tema "Apprentissage de règles à des PLM".
Junto a cada fuente en la lista de referencias hay un botón "Agregar a la bibliografía". Pulsa este botón, y generaremos automáticamente la referencia bibliográfica para la obra elegida en el estilo de cita que necesites: APA, MLA, Harvard, Vancouver, Chicago, etc.
También puede descargar el texto completo de la publicación académica en formato pdf y leer en línea su resumen siempre que esté disponible en los metadatos.
Explore tesis sobre una amplia variedad de disciplinas y organice su bibliografía correctamente.
Ahmadi, Naser. "A framework for the continuous curation of a knowledge base system". Electronic Thesis or Diss., Sorbonne université, 2021. http://www.theses.fr/2021SORUS320.
Texto completoEntity-centric knowledge graphs (KGs) are becoming increasingly popular for gathering information about entities. The schemas of KGs are semantically rich, with many different types and predicates to define the entities and their relationships. These KGs contain knowledge that requires understanding of the KG’s structure and patterns to be exploited. Their rich data structure can express entities with semantic types and relationships, oftentimes domain-specific, that must be made explicit and understood to get the most out of the data. Although different applications can benefit from such rich structure, this comes at a price. A significant challenge with KGs is the quality of their data. Without high-quality data, the applications cannot use the KG. However, as a result of the automatic creation and update of KGs, there are a lot of noisy and inconsistent data in them and, because of the large number of triples in a KG, manual validation is impossible. In this thesis, we present different tools that can be utilized in the process of continuous creation and curation of KGs. We first present an approach designed to create a KG in the accounting field by matching entities. We then introduce methods for the continuous curation of KGs. We present an algorithm for conditional rule mining and apply it on large graphs. Next, we describe RuleHub, an extensible corpus of rules for public KGs which provides functionalities for the archival and the retrieval of rules. We also report methods for using logical rules in two different applications: teaching soft rules to pre-trained language models (RuleBert) and explainable fact checking (ExpClaim)
Ducellier, Guillaume. "Gestion de règles expertes en ingénierie collaborative : applications aux plateformes PLM". Troyes, 2008. http://www.theses.fr/2008TROY0008.
Texto completoProduct Lifecycle Management (PLM) facilitates the creation, change and sharing of product data through the various phases of its definition. However, PLM platforms are based on IT support enabling the data exchange rather than the exchange of information. The proposed approach aims to enhance PLM platform functionalities via interactive product information exchanges (parameters and rules). Parameters and rules are largely used within product data for specifying configurable product. The proposal is to develop some facilities for parameters and rules management integrated to PLM platform for improving the information exchange. The parameters set in the PLM platforms are used by the designer to create product data. The approach enables the definition of parameter sets that describe relevant information on the product definition within the PLM platform. Rules created result from the relations between parameters. Finally, links associate product data that use the same parameters. This tends to formalize relations between parameters and product data in a PLM. In order to enable the implementation of the proposed approach, scenarios have been specified to clarify the various required functionalities. Based on these scenarios, an IT-demonstrator integrated to a PLM platform has been developed. The final results are discussed through the definition of a study case and perspectives are presented
Bannour, Sondes. "Apprentissage interactif de règles d'extraction d'information textuelle". Thesis, Sorbonne Paris Cité, 2015. http://www.theses.fr/2015USPCD113/document.
Texto completoNon communiqué
Chevaleyre, Yann. "Apprentissage de règles à partir de données multi-instances". Paris 6, 2001. http://www.theses.fr/2001PA066502.
Texto completoNakoula, Yassar. "Apprentissage des modèles linguistiques flous, par jeu de règles pondérées". Chambéry, 1997. http://www.theses.fr/1997CHAMS018.
Texto completoHoarau, Martine. "Apprentissage implicite et vieillissement : étude de l'acquisition incidente de règles". Montpellier 3, 2003. http://www.theses.fr/2003MON30003.
Texto completoIn this study, we compared old and young subjects performance in implicit and explicit learning conditions. We presented tasks which differ in the nature of the material (verbal vs visuo-spatial) and the level of processing (associative vs inferential) required to perform the tasks. First, the hypothesis of the reduction of attentional resources with age suggest that the elderly subjects (i) would perform better in implicit than in explicit learning condition and (ii) would perform better the associative than the inferential tasks. Second, the hypothesis of a decline of visuospatial abilities and a preservation of verbal abilities with age suggest that the elderly would perform better the verbal than the visuospatial tasks. The results indicate the effects of the nature of the material and of the level of processing, but no effect of the learning condition. Theses results suggest that aging leads both to the deficit of global resources and to the decline of specific type of processing
Monfardini, Elisabetta. "L' apprentissage social de règles chez l'homme et le singe". Aix-Marseille 2, 2009. http://www.theses.fr/2009AIX22026.
Texto completoMost of our everyday behaviours are guided by conventional rules (‘red traffic light means stop'). These rules, based on arbitrary associations between stimuli and actions, must be learned. We can learn them individually, by trial and error, or, at a lower cost, via observation of others. The social transmission of rules, despite its importance for any gregarious animal species, has seldom been studied and it neural bases remain unexplored. My thesis inaugurates a neuroscience approach of social rule learning. Two main questions are addressed: 1. Is social rule learning based on the same neural bases as individual rule learning in humans? 2. Can the rhesus macaque, a crucial model for neuroscience but considered as a poor imitator by comparative psychologists, be a suitable animal model to explore how the brain learns from others? The first part of my experimental contribution includes two fMRI studies on healthy human volunteers. It demonstrates the existence of a common brain network for the recall of rules, whether learned by trial and error or by observation, as well as similar brain activation dynamics during individual and social learning. The second part consists of three behavioural studies in rhesus monkeys. It confirms the validity of this animal model regarding imitation by revealing a) a spontaneous tendency of this species to observe and take advantage of a conspecific struggling with new rules, b) an interesting propensity to learn more from others' errors than from its own, and c) an ability to learn not only from conspecifics but also from humans
Burg, Bernard. "Apprentissage de règles de comportement destinées au contrôle d'un système". Paris 11, 1988. http://www.theses.fr/1988PA112375.
Texto completoProcess control systems have to face applications which are always more ambitions and difficult to master. In some cases it is not easy to use conventional process control techniques. With the introduction of declarative methods it is possible to start in a pragmatic way and to set an implicit formulation of the problem when no explicit formulation is available. New mechanisms can be envisioned, and we conceived a rule based controller, then the difficulty remains on the design of the rule sets. To overcome this problem, we had to use jointly some learning techniques, such as data analysis to cope with noisy data and to project them into reduced space representations. Then structural techniques allow to modelise the temporal evolution of the process control and the hidden structures. Finally, artificial intelligence machine learning techniques discover the concepts and generalise the acquired knowledge. The whole technique set is supervised by artificial intelligence, it analyses the results issued from each learning step and planes the next action to perform. Three learning strategies are used: the first one starts from the data and uses inductive learning, it proves some completeness. The second one begins with a fuzzy model and acquires rules by deduction, it brings coherency via expert knowledge. Finally the behavior rules are used and refined by means of interaction with the environment. The learning program CANDIDE performed two case studies - the speed control of a DC motor the automatic driving of a car
Guillaume, Serge. "Induction de règles floues interprétables". Toulouse, INSA, 2001. http://www.theses.fr/2001ISAT0021.
Texto completoThis report deals with interpretable fuzzy rule induction from data for human-computer cooperation purposes. A review of fuzzy rule induction methods shows that they can be grouped into three families. Their comparison highlights the fact that the interpretability is not guaranteed. The central part of our work is a new fuzzy rule induction method. It aims to fulfill three interpretability conditions: readable fuzzy partitions, a number of rules as small as possible, incomplete rules. This is achieved through a three step procedure: generating a family of fuzzy partitions for each input variable, building an accurate fuzzy inference system, simplifying the rule base. The procedure is based on original concepts such as a metric distance suitable for fuzzy partitioning, and the input context defined by a set of rules. We introduced coverage and heterogeneity related indices to guide the prodedure, complementary with a numerical performance index. The method is first validated using well known data and then applied to decison making in a complex system. This application means to extract winemaking rules which enhance the color of red wine
Guillame-bert, Mathieu. "Apprentissage de règles associatives temporelles pour les séquences temporelles de symboles". Phd thesis, Université de Grenoble, 2012. http://tel.archives-ouvertes.fr/tel-00849087.
Texto completoGuillame-Bert, Mathieu. "Apprentissage de règles associatives temporelles pour les séquences temporelles de symboles". Thesis, Grenoble, 2012. http://www.theses.fr/2012GRENM081/document.
Texto completoThe learning of temporal patterns is a major challenge of Data mining. We introduce a temporal pattern model called Temporal Interval Tree Association Rules (Tita rules or Titar). This pattern model can be used to express both uncertainty and temporal inaccuracy of temporal events. Among other things, Tita rules can express the usual time point operators, synchronicity, order, and chaining,disjunctive time constraints, as well as temporal negation. Tita rules are designed to allow predictions with optimum temporal precision. Using this representation, we present the Titar learner algorithm that can be used to extract Tita rules from large datasets expressed as Symbolic Time Sequences. This algorithm based on entropy minimization, apriori pruning and statistical dependence analysis. We evaluate our technique on simulated and real world datasets. The problem of temporal planning with Tita rules is studied. We use Tita rules as world description models for a Planning and Scheduling task. We present an efficient temporal planning algorithm able to deal with uncertainty, temporal inaccuracy, discontinuous (or disjunctive) time constraints and predictable but imprecisely time located exogenous events. We evaluate our technique by joining a learning algorithm and our planning algorithm into a simple reactive cognitive architecture that we apply to control a robot in a virtual world
Havard, Christelle. "Entreprise, efficacité et règles organisationnelles : analyse de la cohérence et de la pertinence des règles de deux établissements postaux". Rennes 2, 2000. http://www.theses.fr/2000REN20031.
Texto completoThis thesis deals with changes in company rules and how the effect organisationall efficiency. The hypothesis is that organizational efficiency can be reached only when the rules are both consistent and relevant. A grid was designed for reading and analysing company rules. This grid was applied to two sorting offices and the organisational efficiency of the two establishments subsequently examined
Braud, Raphaël. "Apprentissage incrémental de règles sensorimotrices dans un robot, du babillage moteur à l'utilisation d'outils". Thesis, Cergy-Pontoise, 2017. http://www.theses.fr/2017CERG0897/document.
Texto completoInspired by concepts found in developmental psychology, my work focuses on robotic learning through motor babbling in order to achieve low-level sensorimotor control and, subsequently, to progress to more high-level behaviours such as the use of tools. Tool-use raises several key issues related to the extension of the body schema and the ability to make sequences of actions. In this presentation I will discuss my research efforts in this area by presenting a model called "Dynamic Sensorimotor Model (DSM)". DSM learns sensorimotor laws by making predictions about sensory input variations, as a result of observing environmental phenomena and interacting with objects in the reaching space.Sensorimotor laws depend on; 1) motor magnitudes (e.g., motor commands in velocity) and 2) a given context (i.e., a sensory input vector). A predictor learns and refines sensorimotor laws either during the execution of a task or during a motor babbling phase. Learning laws is therefore independent of the execution of specific tasks and they can be exploited in both new contexts and/or for new tasks.DSM employs two mechanisms. First, a mechanism for motor simulations that considers the result of simulated motor inputs to determine appropriate motor commands to be performed towards a particular task. Second, a mechanism for context simulations that uses simulated sensory inputs in order to identify contexts that can potentially form sub-goals towards the completion of a task.The performance of the system is evaluated through a series of experiments conducted using both a simulated and a real robotic platform. The results demonstrate the ability of the system to complete reaching tasks and highlight its strength in making use of a nearby tool when the target is not within its reach.The ability to make sequences of actions on the fly is based on the accuracy of the contexts that the system gradually learns. The last part of my work focuses on improving the efficiency of making sequences of actions by offering the ability to categorize contexts based on the variations observed in the sensors with respect to the variation of the sensorimotor laws
Guarda, Alvaro. "Apprentissage génétique de règles de reconnaissance visuelle : application à la reconnaissance d'éléments du visage". Grenoble INPG, 1998. http://www.theses.fr/1998INPG0110.
Texto completoD'Ambrosio, Roberto. "Classification de bases de données déséquilibrées par des règles de décomposition". Thesis, Nice, 2014. http://www.theses.fr/2014NICE4007/document.
Texto completoDisproportion among class priors is encountered in a large number of domains making conventional learning algorithms less effective in predicting samples belonging to the minority classes. We aim at developing a reconstruction rule suited to multiclass skewed data. In performing this task we use the classification reliability that conveys useful information on the goodness of classification acts. In the framework of One-per-Class decomposition scheme we design a novel reconstruction rule, Reconstruction Rule by Selection, which uses classifiers reliabilities, crisp labels and a-priori distributions to compute the final decision. Tests show that system performance improves using this rule rather than using well-established reconstruction rules. We investigate also the rules in the Error Correcting Output Code (ECOC) decomposition framework. Inspired by a statistical reconstruction rule designed for the One-per-Class and Pair-Wise Coupling decomposition approaches, we have developed a rule that applies softmax regression on reliability outputs in order to estimate the final classification. Results show that this choice improves the performances with respect to the existing statistical rule and to well-established reconstruction rules. On the topic of reliability estimation we notice that small attention has been given to efficient posteriors estimation in the boosting framework. On this reason we develop an efficient posteriors estimator by boosting Nearest Neighbors. Using Universal Nearest Neighbours classifier we prove that a sub-class of surrogate losses exists, whose minimization brings simple and statistically efficient estimators for Bayes posteriors
Cleuziou, Guillaume. "Une méthode de classification non-supervisée pour l'apprentissage de règles et la recherche d'information". Phd thesis, Université d'Orléans, 2004. http://tel.archives-ouvertes.fr/tel-00084828.
Texto completoNous proposons, dans cette étude, l'algorithme de clustering PoBOC permettant de structurer un ensemble d'objets en classes non-disjointes. Nous utilisons cette méthode de clustering comme outil de traitement dans deux applications très différentes.
- En apprentissage supervisé, l'organisation préalable des instances apporte une connaissance utile pour la tâche d'induction de règles propositionnelles et logiques.
- En Recherche d'Information, les ambiguïtés et subtilités de la langue naturelle induisent naturellement des recouvrements entre thématiques.
Dans ces deux domaines de recherche, l'intérêt d'organiser les objets en classes non-disjointes est confirmé par les études expérimentales adaptées.
Guély, François. "Apprentissage de bases de règles floues : contribution à une étude systématique de l'approche de l'optimisation". Châtenay-Malabry, Ecole centrale de Paris, 1994. http://www.theses.fr/1994ECAP0387.
Texto completoPomorski, Denis. "Apprentissage automatique symbolique/numérique : construction et évaluation d'un ensemble de règles à partir des données". Lille 1, 1991. http://www.theses.fr/1991LIL10117.
Texto completoJrad, Nisrine. "Apprentissage et qualification des règles de décision multiclasses avec rejet sélectif et contraintes de performance". Troyes, 2009. http://www.theses.fr/2009TROY0023.
Texto completoThis thesis deals with supervised learning and quality assessment of a decision rule for multiclass problems with class-selective rejection and performance constraints. The learning process consists in solving and optimisation problem with constraints for a given model. Several families of decision rules with different complexity can be defined by restricting parameter domain. An optimal rule can be obtained within each family. In order to compare the different rules and chose the best one, a common criterion should be defined. This criterion should take into consideration the constraints. Thus, this criterion if function of the decision rule and the weights associated to each constraint. A performance criterion is proposed and its estimation is discussed. Two learning processes, givent by a class-modelling approache and a boundary approache, based on one class SVM, are presented. Synthetic complementary problems are also tackled such as learning decision rule for problems with time evoluarionary constraints or using a cascade classifier system with class-selective rejection to improve the accuracy of the decision. This latter study was applied to cancer tumours diagnosis and results in significant performance improvement
Salvat, Eric. "Raisonner avec des opérations de graphes : graphes conceptuels et règles d'inférence". Montpellier 2, 1997. http://www.theses.fr/1997MON20192.
Texto completoD'ambrosio, Roberto. "Classification de bases de données déséquilibrées par des règles de décomposition". Phd thesis, Université Nice Sophia Antipolis, 2014. http://tel.archives-ouvertes.fr/tel-00995021.
Texto completoChraibi, Kaadoud Ikram. "apprentissage de séquences et extraction de règles de réseaux récurrents : application au traçage de schémas techniques". Thesis, Bordeaux, 2018. http://www.theses.fr/2018BORD0032/document.
Texto completoThere are two important aspects of the knowledge that an individual acquires through experience. One corresponds to the semantic memory (explicit knowledge, such as the learning of concepts and categories describing the objects of the world) and the other, the procedural or syntactic memory (knowledge relating to the learning of rules or syntax). This "syntactic memory" is built from experience and particularly from the observation of sequences of objects whose organization obeys syntactic rules.It must have the capability to aid recognizing as well as generating valid sequences in the future, i.e., sequences respecting the learnt rules. This production of valid sequences can be done either in an explicit way, that is, by evoking the underlying rules, or implicitly, when the learning phase has made it possible to capture the principle of organization of the sequences without explicit recourse to the rules. Although the latter is faster, more robust and less expensive in terms of cognitive load as compared to explicit reasoning, the implicit process has the disadvantage of not giving access to the rules and thus becoming less flexible and less explicable. These mnemonic mechanisms can also be applied to business expertise. The capitalization of information and knowledge in general, for any company is a major issue and concerns both the explicit and implicit knowledge. At first, the expert makes a choice to explicitly follow the rules of the trade. But then, by dint of repetition, the choice is made automatically, without explicit evocation of the underlying rules. This change in encoding rules in an individual in general and particularly in a business expert can be problematic when it is necessary to explain or transmit his or her knowledge. Indeed, if the business concepts can be formalized, it is usually in any other way for the expertise which is more difficult to extract and transmit.In our work, we endeavor to observe sequences of electrical components and in particular the problem of extracting rules hidden in these sequences, which are an important aspect of the extraction of business expertise from technical drawings. We place ourselves in the connectionist domain, and we have particularly considered neuronal models capable of processing sequences. We implemented two recurrent neural networks: the Elman model and a model with LSTM (Long Short Term Memory) units. We have evaluated these two models on different artificial grammars (Reber's grammar and its variations) in terms of learning, their generalization abilities and their management of sequential dependencies. Finally, we have also shown that it is possible to extract the encoded rules (from the sequences) in the recurrent network with LSTM units, in the form of an automaton. The electrical domain is particularly relevant for this problem. It is more constrained with a limited combinatorics than the planning of tasks in general cases like navigation for example, which could constitute a perspective of this work
Cousin, Marie-Paule. "Apprentissage de la production écrite de l'accord en nombre : application de règles et/ou récuperation d'instances ?" Rouen, 2004. http://www.theses.fr/2004ROUEL477.
Texto completoThis thesis studies the processes involved in the acquisition of the written number flexional morphology by French children. Six experiments are submitted. The first five ones use a cross method. They show the item's previous encounter and the number form impact on number agreement production. The sixth one is a longitudinal study. It examines the impact of the encounter frequency in time, the number form and the age of acquisition. This research shows that children use early instances retrieval for a written number agreement. As for adults, the number flexional morphology processing can be produced by children. In two ways. The first one is the implementation of rules, the second one is the instances direct retrieval
Ndiaye, Seydina Moussa. "Apprentissage par renforcement en horizon fini : Application à la génération de règles pour la conduite de culture". Toulouse 3, 1999. http://www.theses.fr/1999TOU30010.
Texto completoCollain, Emmanuel y Jean-Marc Fovet. "Apprentissage de plans de résolution pour améliorer l'efficacité des chainages avant des systèmes à base de règles". Paris 6, 1991. http://www.theses.fr/1991PA066446.
Texto completoWang, Olivier. "Adaptive Rules Model : Statistical Learning for Rule-Based Systems". Thesis, Université Paris-Saclay (ComUE), 2017. http://www.theses.fr/2017SACLX037/document.
Texto completoBusiness Rules (BRs) are a commonly used tool in industry for the automation of repetitive decisions. The emerging problem of adapting existing sets of BRs to an ever-changing environment is the motivation for this thesis. Existing Supervised Machine Learning techniques can be used when the adaptation is done knowing in detail which is the correct decision for each circumstance. However, there is currently no algorithm, theoretical or practical, which can solve this problem when the known information is statistical in nature, as is the case for a bank wishing to control the proportion of loan requests its automated decision service forwards to human experts. We study the specific learning problem where the aim is to adjust the BRs so that the decisions are close to a given average value.To do so, we consider sets of Business Rules as programs. After formalizing some definitions and notations in Chapter 2, the BR programming language defined this way is studied in Chapter 3, which proves that there exists no algorithm to learn Business Rules with a statistical goal in the general case. We then restrain the scope to two common cases where BRs are limited in some way: the Iteration Bounded case in which no matter the input, the number of rules executed when taking the decision is less than a given bound; and the Linear Iteration Bounded case in which rules are also all written in Linear form. In those two cases, we later produce a learning algorithm based on Mathematical Programming which can solve this problem. We briefly extend this theory and algorithm to other statistical goal learning problems in Chapter 5, before presenting the experimental results of this thesis in Chapter 6. The last includes a proof of concept to automate the main part of the learning algorithm which does not consist in solving a Mathematical Programming problem, as well as some experimental evidence of the computational complexity of the algorithm
Averyanov, Yaroslav. "Concevoir et analyser de nouvelles règles d’arrêt prématuré pour économiser les ressources de calcul". Thesis, Lille 1, 2020. http://www.theses.fr/2020LIL1I048.
Texto completoThis work develops and analyzes strategies for constructing instances of the so-called early stopping rules applied to some iterative learning algorithms for estimating the regression function. Such quantities are data-driven rules indicating when to stop the iterative learning process to reach a trade-off between computational costs and the statistical precision. Unlike a large part of the existing literature on early stopping, where these rules only depend on the data in a "weak manner", we provide data-driven solutions for the aforementioned problem without utilizing validation data. The crucial idea exploited here is that of the minimum discrepancy principle (MDP), which shows when to stop an iterative learning algorithm. To the best of our knowledge, this idea dates back to the work of Vladimir A. Morozov in the 1960s-1970s who studied linear ill-posed problems and their regularization, mostly inspired by mathematical physics problems. Among different applications of this line of work, the so-called spectral filter estimators such as spectral cut-off, Landweber iterations, and Tikhonov (ridge) regularization have received quite a lot of attention (e.g., in statistical inverse problems). It is worth mentioning that the minimum discrepancy principle consists in controlling the residuals of an estimator (which are iteratively minimized) and properly setting a threshold for them such that one can achieve some (minimax) optimality. The first part of this thesis is dedicated to theoretical guarantees of stopping rules based on the minimum discrepancy principle and applied to gradient descent, and Tikhonov (ridge) regression in the framework of reproducing kernel Hilbert space (RKHS). There, we show that this principle provides a minimax optimal functional estimator of the regression function when the rank of the kernel is finite. However, when one deals with infinite-rank reproducing kernels, the resulting estimator will be only suboptimal. While looking for a solution, we found the existence of the so-called residuals polynomial smoothing strategy. This strategy (combined with MDP) has been proved to be optimal for the spectral cut-off estimator in the linear Gaussian sequence model. We borrow this strategy, modify the stopping rule accordingly, and prove that the smoothed minimum discrepancy principle yields a minimax optimal functional estimator over a range of function spaces, which includes the well-known Sobolev function class. Our second contribution consists in exploring the theoretical properties of the minimum discrepancy stopping rule applied to the more general family of linear estimators. The main difficulty of this approach is that, unlike the spectral filter estimators considered earlier, linear estimators do no longer lead to monotonic quantities (the bias and variance terms). Let us mention that this is also the case for famous algorithms such as Stochastic Gradient Descent. Motivated by further practical applications, we work with the widely used k-NN regression estimator as a reliable first example. We prove that the aforementioned stopping rule leads to a minimax optimal functional estimator, in particular, over the class of Lipschitz functions on a bounded domain.The third contribution consists in illustrating through empirical experiments that for choosing the tuning parameter in a linear estimator (the k-NN regression, Nadaraya-Watson, and variable selection estimators), the MDP-based early stopping rule performs comparably well with respect to other widely used and known model selection criteria
Beaudoin, Mathieu. "Découverte de règles de classification pour un système d’aide à la décision pour la surveillance de l'usage des antimicrobiens". Thèse, Université de Sherbrooke, 2015. http://hdl.handle.net/11143/7591.
Texto completoKabdebon, Claire. "Sequence encoding in preverbal infants : an electrophysiological perspective". Thesis, Paris 6, 2016. http://www.theses.fr/2016PA066614.
Texto completoTo this day, the infant brain is the only known learning system able to apprehend and master the complexity of the human language. Developmental psychologists have dedicated a lot of efforts to break down the mystery of language acquisition, revealing precocious and impressive abilities for processing and encoding speech sequences. The recent emergence of non-invasive neuroimaging techniques provides a new tool to explore language learning mechanisms from a different perspective. In the present thesis, we aimed at investigating the encoding mechanisms of the structural properties of a speech sequence from an electrophysiological perspective. In the first part of this thesis, we provided the developmental neuroimaging community with a methodological contribution. Based on magnetic resonance imaging (MRI) data, we virtually localized the standardized sensor placement system for both electroencephalography (EEG) and near infrared spectroscopy (NIRS) relative to the internal brain structures, and assessed their variability. We additionally provided an infant brain template with an anatomical atlas which will be valuable for studies in which individual anatomical information cannot be obtained. In the second part of this thesis, using high-density EEG, we demonstrated that 8 month-old infants could deploy powerful learning mechanisms for capturing the statistical dependencies between non-adjacent syllable units, in order to chunk a continuous speech stream. Interestingly, a hierarchy of neural processes tracked both the syllables and the chunked constituents of the sequence. Finally, in a third cognitive EEG study, we proposed an experimental design to assess infants’ ability to not only extract but also encode the structure of speech sequences into unified mental schemas. The results of this study established that 5 month-old infants could form robust mental representations for repetition-based sequences, allowing them to represent, categorize and operate on multiple structures. Inspection of various neural measurements revealed that several stages of the processing hierarchy were affected by the acquired mental representations. Overall, this thesis complements behavioral research on language acquisition with a window onto the early neural mechanisms allowing sequence encoding, revealing a hierarchy of increasingly complex computations in the encoding of linguistic structures
Perrin, Pierre. "Territoires et diffusion des règles sociales : vers une théorie de la coexistence et de la convergence institutionnelles". Aix-Marseille 3, 2003. http://www.theses.fr/2003AIX32022.
Texto completoAn issue at stake in institutional competition between territories is to determine to what it leads: institutional standardization or diversity?Several models answer to this question. However, their fundamental assumptions lead to ignore some important characteristics of the diffusion process. This work tries to remedy to this limit. Through an analysis of proximity into social networks, this work shows that rules transform when they spread within territories. Secondly, when the identity of territories is perceived as a set of interdependent rules, the dissertation explains how the differentiation of rules that spreads, aims at making them consistent with the institutions of the different territories. Then, the diffusion of an institution does not lead to an institutional and territorial standardization. This process induces the persistence of institutional competition between territories
Benard, Julie. "Apprentissages visuels chez l'abeille Apis mellifera : de la généralisation à l'extraction de règles". Toulouse 3, 2007. http://thesesups.ups-tlse.fr/125/.
Texto completoThis work presents different kind of visual learning with increasing cognitive requirements in free-flying honeybees Apis mellifera, from color generalization to rules extraction. We first study color generalization in bees trained to two rewarding colors, and we observe that their performances are consistent with a linear summation of the two generalization gradients generated by two trained colors. These gradients are asymmetric as control bees respond to the test stimuli as if these belong to different similarity classes in spite of having similar perceptual distances separating them. Our results suggest that color categories could exist in honeybees. Therefore, in a second study we show that bees can master a color categorization task. Bees treat bluish and yellowish stimuli as belonging to different classes. In a third study, we show that bees trained with complex achromatic patterns sharing a common layout comprising four edge orientations remember these orientations simultaneously in their appropriate positions. Moreover, we show that bees can use this kind of representation to categorize novel stimuli, and that stimulation of the achromatic L-photoreceptor is necessary for this task. In the last study, we asked whether bees could solve a transitive inference problem. We find that bees do not establish transitive inferences but, rather, guide their choices by the joint action of a recency effect and the associative strength of the stimuli. This all work contributes to a better comprehension of the visual cognitive abilities in insects and to determine precisely the abilities of the small bee brain to situate cognitive research into an appropriate comparative frame
Abramé, André. "Max-résolution et apprentissage pour la résolution du problème de satisfiabilité maximum". Thesis, Aix-Marseille, 2015. http://www.theses.fr/2015AIXM4330/document.
Texto completoThis PhD thesis is about solving the Maximum Satisfiability (Max-SAT) problem. We study the mechanisms related to the detection and transformations of the inconsistent subsets by the max-resolution rule. In the context of the branch and bound (BnB) algorithms, we present several contributions related to the lower bound computation. They range from the study of the unit propagation scheme used to detect inconsistent subsets to the extension of the learning criteria and to the evaluation of the impact of the max-resolution transformations on the BnB solvers efficiency. Thanks to our contributions, we have implemented a new solver which is competitive with the state of art ones. We give insights allowing a better understanding of the behavior of BnB solvers as well as theoretical elements which contribute to explain the efficiency of these solvers and their limits. It opens new development perspectives on the learning mechanisms used by BnB solvers which may lead to a better consideration of the instances structural properties. We also present an example of integration of the max-resolution inference rule in a local search algorithm
Rossant, Florence. "Reconnaissance de partitions musicales par modélisation floue des informations extraites et des règles de notation". Phd thesis, Télécom ParisTech, 2006. http://pastel.archives-ouvertes.fr/pastel-00002037.
Texto completoBelhaoues, Rachid. "La découverte de règles chez l'homme et la souris : étude chez des sujets sains et cérébro-lésés". Aix-Marseille 1, 2004. http://www.theses.fr/2004AIX11022.
Texto completoSi, Fodil David Mohand. "Commande floue et optimisation de base de règles floues pour la régulation de la réactivité et de la température moyenne dans les réacteurs nucléaires à eau pressurisée". Châtenay-Malabry, Ecole centrale de Paris, 2000. http://www.theses.fr/2000ECAP0865.
Texto completoAudet, Marie-Josée. "Apprentissage et critique de règles de la communication publique lors d'un débat public : le cas des représentants étudiants lors de la grève de 2012 au Québec". Master's thesis, Université Laval, 2017. http://hdl.handle.net/20.500.11794/27536.
Texto completoThis thesis deals with the methods of learning and criticizing the rules of public communication by student representatives during the student strike in 2012 in Quebec. Taking into account the particular characteristics of student organizations and the conditions in which student representatives learn the rudiments of the roles of press officer and spokesperson, this paper proposes to study the modalities of learning and criticism of rules of public communication. New actors who do not benefit from the same human and financial resources as government and media actors, but with whom they publicly debated during the 2012 student strike. How did they learn the rules of public communication in an intense and animated public debate which subsequently degenerated into a social crisis? As a result of the 2012 experience, these individuals became much more known and recognized, and gained notoriety. We want to understand how the different ways of learning about the rules of public communication depend on the particular environment and the context in which the student representatives were at that time. We also want to focus on the content of learning and criticism to better understand what kinds of rules are learned and what are their strategic functions. The research is based on a qualitative interview analysis of the public affair program 24 heures en 60 minutes, on RDI, and a series of semi-conducted interviews with student representatives. Keywords: Criticism of rules of public communication; Learning; New players; Public debate; Student strike of 2012
Ghemmogne, Fossi Leopold. "Gestion des règles basée sur l'indice de puissance pour la détection de fraude : Approches supervisées et semi-supervisées". Thesis, Lyon, 2019. http://www.theses.fr/2019LYSEI079.
Texto completoThis thesis deals with the detection of credit card fraud. According to the European Central Bank, the value of frauds using cards in 2016 amounted to 1.8 billion euros. The challenge for institutions is to reduce these frauds. In general, fraud detection systems consist of an automatic system built with "if-then" rules that control all incoming transactions and trigger an alert if the transaction is considered suspicious. An expert group checks the alert and decides whether it is true or not. The criteria used in the selection of the rules that are kept operational are mainly based on the individual performance of the rules. This approach ignores the non-additivity of the rules. We propose a new approach using power indices. This approach assigns to the rules a normalized score that quantifies the influence of the rule on the overall performance of the group. The indexes we use are the Shapley Value and Banzhaf Value. Their applications are 1) Decision support to keep or delete a rule; 2) Selection of the number k of best-ranked rules, in order to work with a more compact set. Using real credit card fraud data, we show that: 1) This approach performs better than the one that evaluates the rules in isolation. 2) The performance of the set of rules can be achieved by keeping one-tenth of the rules. We observe that this application can be considered as a task of selection of characteristics: We show that our approach is comparable to the current algorithms of the selection of characteristics. It has an advantage in rule management because it assigns a standard score to each rule. This is not the case for most algorithms, which focus only on an overall solution. We propose a new version of Banzhaf Value, namely k-Banzhaf; which outperforms the previous in terms of computing time and has comparable performance. Finally, we implement a self-learning process to reinforce the learning in an automatic learning algorithm. We compare these with our power indices to rank credit card fraud data. In conclusion, we observe that the selection of characteristics based on the power indices has comparable results with the other algorithms in the self-learning process
Mita, Graziano. "Toward interpretable machine learning, with applications to large-scale industrial systems data". Electronic Thesis or Diss., Sorbonne université, 2021. http://www.theses.fr/2021SORUS112.
Texto completoThe contributions presented in this work are two-fold. We first provide a general overview of explanations and interpretable machine learning, making connections with different fields, including sociology, psychology, and philosophy, introducing a taxonomy of popular explainability approaches and evaluation methods. We subsequently focus on rule learning, a specific family of transparent models, and propose a novel rule-based classification approach, based on monotone Boolean function synthesis: LIBRE. LIBRE is an ensemble method that combines the candidate rules learned by multiple bottom-up learners with a simple union, in order to obtain a final intepretable rule set. Our method overcomes most of the limitations of state-of-the-art competitors: it successfully deals with both balanced and imbalanced datasets, efficiently achieving superior performance and higher interpretability in real datasets. Interpretability of data representations constitutes the second broad contribution to this work. We restrict our attention to disentangled representation learning, and, in particular, VAE-based disentanglement methods to automatically learn representations consisting of semantically meaningful features. Recent contributions have demonstrated that disentanglement is impossible in purely unsupervised settings. Nevertheless, incorporating inductive biases on models and data may overcome such limitations. We present a new disentanglement method - IDVAE - with theoretical guarantees on disentanglement, deriving from the employment of an optimal exponential factorized prior, conditionally dependent on auxiliary variables complementing input observations. We additionally propose a semi-supervised version of our method. Our experimental campaign on well-established datasets in the literature shows that IDVAE often beats its competitors according to several disentanglement metrics
Cadot, Martine. "Extraire et valider les relations complexes en sciences humaines : statistiques, motifs et règles d'association". Phd thesis, Université de Franche-Comté, 2006. http://tel.archives-ouvertes.fr/tel-00594174.
Texto completoTsopze, Norbert. "Treillis de Galois et réseaux de neurones : une approche constructive d'architecture des réseaux de neurones". Thesis, Artois, 2010. http://www.theses.fr/2010ARTO0407/document.
Texto completoThe artificial neural networks are successfully applied in many applications. But theusers are confronted with two problems : defining the architecture of the neural network able tosolve their problems and interpreting the network result. Many research works propose some solutionsabout these problems : to find out the architecture of the network, some authors proposeto use the problem domain theory and deduct the network architecture and some others proposeto dynamically add neurons in the existing networks until satisfaction. For the interpretabilityproblem, solutions consist to extract rules which describe the network behaviour after training.The contributions of this thesis concern these problems. The thesis are limited to the use of theartificial neural networks in solving the classification problem.In this thesis, we present a state of art of the existing methods of finding the neural networkarchitecture : we present a theoritical and experimental study of these methods. From this study,we observe some limits : difficulty to use some method when the knowledges are not available ;and the network is seem as ’black box’ when using other methods. We a new method calledCLANN (Concept Lattice-based Artificial Neural Network) which builds from the training dataa semi concepts lattice and translates this semi lattice into the network architecture. As CLANNis limited to the two classes problems, we propose MCLANN which extends CLANN to manyclasses problems.A new method of rules extraction called ’MaxSubsets Approach’ is also presented in thisthesis. Its particularity is the possibility of extracting the two kind of rules (If then and M-of-N)from an internal structure.We describe how to explain the MCLANN built network result aboutsome inputs
Latiri, Chiraz. "Extraction de Connaissances a partir de Textes : M ethodes et Applications". Habilitation à diriger des recherches, Université de Lorraine, 2013. http://tel.archives-ouvertes.fr/tel-00927238.
Texto completoDaraut, Sandrine. "De l'apprentissage technico-organisationnel ou du rôle des règles dans la structuration de contextes d'interactions - Fondements théoriques et illustrations empiriques". Phd thesis, Université des Sciences Sociales - Toulouse I, 2004. http://tel.archives-ouvertes.fr/tel-00141424.
Texto completoNouvel, Damien. "Reconnaissance des entités nommées par exploration de règles d'annotation - Interpréter les marqueurs d'annotation comme instructions de structuration locale". Phd thesis, Université François Rabelais - Tours, 2012. http://tel.archives-ouvertes.fr/tel-00788630.
Texto completoZumpe, Martin Kai. "Stabilité macroéconomique, apprentissage et politique monétaire : une approche comparative : modélisation DSGE versus modélisation multi-agents". Thesis, Bordeaux 4, 2012. http://www.theses.fr/2012BOR40022/document.
Texto completoThis thesis analyses the role of learning in two different modelling frameworks. In the new canonicalmodel with adaptive learning, the most remarkable characteristics of the learning dynamics deal withthe capacity of monetary policy rules to guaranty convergence to the rational expectations equilibrium.The transmission mechanism of the monetary policy is based on the substitution effect associated to theconsumption channel. In the case of an agent-based model which relaxes some restrictive assumptionsof the new canonical model - but is endowed with a similar structure - aggregate variables evolve atsome distance from the rational expectations equilibrium. Monetary policy has a marginal impact onthe agregated variables via the wealth effect of the consumption channel. When agents learn accordingto an evolutionnary social learning process, the economy converges to regions of low economic activity.The introduction of a process where agents learn individually by using their mental models induces lessdepressive learning dynamics. These differences between the two modelling frameworks show that thegeneralisation of the results of the new canonical model is not easy to achieve
Pasquali, Antoine. "Learning with and without consciousness: empirical and computational explorations". Doctoral thesis, Universite Libre de Bruxelles, 2009. http://hdl.handle.net/2013/ULB-DIPOT:oai:dipot.ulb.ac.be:2013/210269.
Texto completoHere are a few of the many questions that I have attempted to investigate during the past few years. The main goal of this thesis was to explore the differences between conscious and unconscious learning. Thus, I will expose the behavioral and computational explorations that we conducted during the last few years. To present them properly, I first review the main concepts that, for almost a century now, researchers in the fields of neuroscience have formulated in order to tackle the issues of both learning and consciousness. Then I detail different hypotheses that guided our empirical and computational explorations. Notably, a few series of experiments allowed identification of several mechanisms that participate in either unconscious or conscious learning. In addition we explored a computational framework for explaining how one could learn unconsciously and nonetheless gain subjective access to one’s mental events. After reviewing the unfolding of our investigation, I detail the mechanisms that we identified as responsible for differences between learning with and without consciousness, and propose new hypotheses to be evaluated in the future.
Doctorat en Sciences Psychologiques et de l'éducation
info:eu-repo/semantics/nonPublished
Colombet-Madinier, Isabelle. "Aspects méthodologiques de la prédiction du risque cardiovasculaire : apports de l'apprentissage automatique". Paris 6, 2002. http://www.theses.fr/2002PA066083.
Texto completoCayla, David. "L'apprentissage organisationnel entre processus adaptatif et changement dirigé". Phd thesis, Université Panthéon-Sorbonne - Paris I, 2007. http://tel.archives-ouvertes.fr/tel-00198591.
Texto completoles conceptions de la rationalité qui existent en économie. Nous pouvons alors établir, en nous appuyant sur l'apport des sciences cognitives contemporaines, une représentation qui permet d'appréhender différents niveaux d'apprentissage, imbriqués et hiérarchisés.
La seconde partie de cette thèse se penche sur l'apprentissage organisationnel de manière plus spécifique. Après avoir montré comment le critère de cohérence pouvait permettre de distinguer les modes de coordination ex post des modes de coordination ex ante, nous nous intéressons au fonctionnement interne des organisations et à la capacité qu'a le management d'en modifier le comportement. Dans le dernier chapitre, enfin, nous montrons quels apports spécifiques notre approche pourrait avoir dans le cadre des théories modernes des
organisations, et nous nous intéressons à la relation entre la structure organisationnelle et la performance de l'apprentissage.
Shahzad, Muhammad Kashif. "Exploitation dynamique des données de production pour améliorer les méthodes DFM dans l'industrie Microélectronique". Phd thesis, Université de Grenoble, 2012. http://tel.archives-ouvertes.fr/tel-00771672.
Texto completoBédard, Olsson Janique. "À la Recherche d'Éléments de Phonétique : Une analyse de la phonétique comme outil de travail dans l'enseignement du français dans les lycées suédois". Thesis, Umeå universitet, Lärarhögskolan vid Umeå universitet (LH), 2012. http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-51597.
Texto completoBandon, David. "Maria : une optimisation adaptative d'un archivage d'images médicales par transfert anticipé". Compiègne, 1996. http://www.theses.fr/1996COMPD961.
Texto completo