Dissertations / Theses on the topic 'Méthodes d’apprentissage en ligne'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Méthodes d’apprentissage en ligne.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
Contal, Emile. "Méthodes d’apprentissage statistique pour l’optimisation globale." Thesis, Université Paris-Saclay (ComUE), 2016. http://www.theses.fr/2016SACLN038/document.
Full textThis dissertation is dedicated to a rigorous analysis of sequential global optimization algorithms. We consider the stochastic bandit model where an agent aim at finding the input of a given system optimizing the output. The function which links the input to the output is not explicit, the agent requests sequentially an oracle to evaluate the output for any input. This function is not supposed to be convex and may display many local optima. In this work we tackle the challenging case where the evaluations are expensive, which requires to design a careful selection of the input to evaluate. We study two different goals, either to maximize the sum of the rewards received at each iteration, or to maximize the best reward found so far. The present thesis comprises the field of global optimization where the function is a realization from a known stochastic process, and the novel field of optimization by ranking where we only perform function value comparisons. We propose novel algorithms and provide theoretical concepts leading to performance guarantees. We first introduce an optimization strategy for observations received by batch instead of individually. A generic study of local supremum of stochastic processes allows to analyze Bayesian optimization on nonparametric search spaces. In addition, we show that our approach extends to natural non-Gaussian processes. We build connections between active learning and ranking and deduce an optimization algorithm of potentially discontinuous functions
Collet, Timothé. "Méthodes optimistes d’apprentissage actif pour la classification." Thesis, Université de Lorraine, 2016. http://www.theses.fr/2016LORR0084/document.
Full textA Classification problem makes use of a training set consisting of data labeled by an oracle. The larger the training set, the best the performance. However, requesting the oracle may be costly. The goal of Active Learning is thus to minimize the number of requests to the oracle while achieving the best performance. To do so, the data that are presented to the oracle must be carefully selected among a large number of unlabeled instances acquired at no cost. However, the true profitability of labeling a particular instance may not be known perfectly. It can therefore be estimated along with a measure of uncertainty. To Increase the precision on the estimate, we need to label more data. Thus, there is a dilemma between labeling data in order to increase the performance of the classifier or to better know how to select data. This dilemma is well studied in the context of finite budget optimization under the name of exploration versus exploitation dilemma. The most famous solutions make use of the principle of Optimism in the Face of Uncertainty. In this thesis, we show that it is possible to adapt this principle to the active learning problem for classification. Several algorithms have been developed for classifiers of increasing complexity, each one of them using the principle of Optimism in the Face of Uncertainty, and their performances have been empirically evaluated
Collet, Timothé. "Méthodes optimistes d’apprentissage actif pour la classification." Electronic Thesis or Diss., Université de Lorraine, 2016. http://www.theses.fr/2016LORR0084.
Full textA Classification problem makes use of a training set consisting of data labeled by an oracle. The larger the training set, the best the performance. However, requesting the oracle may be costly. The goal of Active Learning is thus to minimize the number of requests to the oracle while achieving the best performance. To do so, the data that are presented to the oracle must be carefully selected among a large number of unlabeled instances acquired at no cost. However, the true profitability of labeling a particular instance may not be known perfectly. It can therefore be estimated along with a measure of uncertainty. To Increase the precision on the estimate, we need to label more data. Thus, there is a dilemma between labeling data in order to increase the performance of the classifier or to better know how to select data. This dilemma is well studied in the context of finite budget optimization under the name of exploration versus exploitation dilemma. The most famous solutions make use of the principle of Optimism in the Face of Uncertainty. In this thesis, we show that it is possible to adapt this principle to the active learning problem for classification. Several algorithms have been developed for classifiers of increasing complexity, each one of them using the principle of Optimism in the Face of Uncertainty, and their performances have been empirically evaluated
Colin, Igor. "Adaptation des méthodes d’apprentissage aux U-statistiques." Thesis, Paris, ENST, 2016. http://www.theses.fr/2016ENST0070/document.
Full textWith the increasing availability of large amounts of data, computational complexity has become a keystone of many machine learning algorithms. Stochastic optimization algorithms and distributed/decentralized methods have been widely studied over the last decade and provide increased scalability for optimizing an empirical risk that is separable in the data sample. Yet, in a wide range of statistical learning problems, the risk is accurately estimated by U-statistics, i.e., functionals of the training data with low variance that take the form of averages over d-tuples. We first tackle the problem of sampling for the empirical risk minimization problem. We show that empirical risks can be replaced by drastically computationally simpler Monte-Carlo estimates based on O(n) terms only, usually referred to as incomplete U-statistics, without damaging the learning rate. We establish uniform deviation results and numerical examples show that such approach surpasses more naive subsampling techniques. We then focus on the decentralized estimation topic, where the data sample is distributed over a connected network. We introduce new synchronous and asynchronous randomized gossip algorithms which simultaneously propagate data across the network and maintain local estimates of the U-statistic of interest. We establish convergence rate bounds with explicit data and network dependent terms. Finally, we deal with the decentralized optimization of functions that depend on pairs of observations. Similarly to the estimation case, we introduce a method based on concurrent local updates and data propagation. Our theoretical analysis reveals that the proposed algorithms preserve the convergence rate of centralized dual averaging up to an additive bias term. Our simulations illustrate the practical interest of our approach
Colin, Igor. "Adaptation des méthodes d’apprentissage aux U-statistiques." Electronic Thesis or Diss., Paris, ENST, 2016. http://www.theses.fr/2016ENST0070.
Full textWith the increasing availability of large amounts of data, computational complexity has become a keystone of many machine learning algorithms. Stochastic optimization algorithms and distributed/decentralized methods have been widely studied over the last decade and provide increased scalability for optimizing an empirical risk that is separable in the data sample. Yet, in a wide range of statistical learning problems, the risk is accurately estimated by U-statistics, i.e., functionals of the training data with low variance that take the form of averages over d-tuples. We first tackle the problem of sampling for the empirical risk minimization problem. We show that empirical risks can be replaced by drastically computationally simpler Monte-Carlo estimates based on O(n) terms only, usually referred to as incomplete U-statistics, without damaging the learning rate. We establish uniform deviation results and numerical examples show that such approach surpasses more naive subsampling techniques. We then focus on the decentralized estimation topic, where the data sample is distributed over a connected network. We introduce new synchronous and asynchronous randomized gossip algorithms which simultaneously propagate data across the network and maintain local estimates of the U-statistic of interest. We establish convergence rate bounds with explicit data and network dependent terms. Finally, we deal with the decentralized optimization of functions that depend on pairs of observations. Similarly to the estimation case, we introduce a method based on concurrent local updates and data propagation. Our theoretical analysis reveals that the proposed algorithms preserve the convergence rate of centralized dual averaging up to an additive bias term. Our simulations illustrate the practical interest of our approach
Tempier, Charlotte. "L'Autorégulation dans un dispositif en ligne d’apprentissage : Signes d’un apprenant opportuniste." Paris 10, 2011. http://www.theses.fr/2011PA100004.
Full textBouaziz, Ameni. "Méthodes d’apprentissage interactif pour la classification des messages courts." Thesis, Université Côte d'Azur (ComUE), 2017. http://www.theses.fr/2017AZUR4039/document.
Full textAutomatic short text classification is more and more used nowadays in various applications like sentiment analysis or spam detection. Short texts like tweets or SMS are more challenging than traditional texts. Therefore, their classification is more difficult owing to their shortness, sparsity and lack of contextual information. We present two new approaches to improve short text classification. Our first approach is "Semantic Forest". The first step of this approach proposes a new enrichment method that uses an external source of enrichment built in advance. The idea is to transform a short text from few words to a larger text containing more information in order to improve its quality before building the classification model. Contrarily to the methods proposed in the literature, the second step of our approach does not use traditional learning algorithm but proposes a new one based on the semantic links among words in the Random Forest classifier. Our second contribution is "IGLM" (Interactive Generic Learning Method). It is a new interactive approach that recursively updates the classification model by considering the new data arriving over time and by leveraging the user intervention to correct misclassified data. An abstraction method is then combined with the update mechanism to improve short text quality. The experiments performed on these two methods show their efficiency and how they outperform traditional algorithms in short text classification. Finally, the last part of the thesis concerns a complete and argued comparative study of the two proposed methods taking into account various criteria such as accuracy, speed, etc
Renaudie, David. "Méthodes d’apprentissage automatique pour la modélisation de l’élève en algèbre." Grenoble INPG, 2005. http://www.theses.fr/2005INPG0008.
Full textBuilding an Intelligent Tutoring System which adapts to student difficulties, needs to develop automatic tools that diagnose their knowledge level. This work is based on a database containing behaviours of students, who are learning algebra with the Aplusix software. We aim at automatically extracting behavioural regularities from this database, in order to support the developpement of an intelligent tutor. To achieve this goal, we apply machine learning methods that detect similarities in datasets, and we propose two distinct approaches to student modelling issues. One one hand, we identify clusters of students that share behavioural similarities on a given exercise, with an unsupervised clustering algorithm. On the other hand, an analysis of procedural regularities in the production of each student, inspired from a theoretical framework of knowledge representation and based on a symbolic learning algorithm, leads to individual models that could be used for adapted remediation
Achoch, Mounia. "Méthodes d’apprentissage et approches expérimentales appliqués aux réseaux d’interfaces protéiques." Thesis, Université Grenoble Alpes (ComUE), 2015. http://www.theses.fr/2015GREAA022/document.
Full textThe aim of this study is to understand protein assembly mechanisms. The assembly of a protein in an oligomer is particularly important because it is involved in many pathologies going from bacterial infection, Alzheimer like diseases or even some cancers. Protein assembly is the combination of two or more protein chains to induce a biological activity. The B subunit of the cholera toxin pentamer (CtxB5), which belongs to the family of AB5 toxins, is studied as the main model of assembly. Experimental results have provided information on the assembly of the toxin highlighting the involvement of certain amino acids. The first problem addressed in my thesis is to understand their role and see if network approaches are relevant to such investigation. I was able to show using amino acid mutations, that amino acids influence each other by cascade or "peer to peer" mechanisms in order to coordinate the various steps of the assembly (Chapters 4, 5 and 6). The structure and function of the proteins are defined by amino acid sequences which naturally vary due to genetic mutation. So I decided to expand this field of investigation to see if the cascade mechanism was generalized as a mean of disrupting a protein structure. Here it is to understand how a protein loses its function by way of a significant change of structure upon mutation. First, I studied dataset to know the characteristics of healthy protein networks (Chapter 7, 8 and 9), and after I looked at the effects of the systematic mutation of each amino acid of CtxB5 on its overall structure (Chapter 10 and 11). Mutations led from moderate to very large structural changes around the mutated amino acid or at long distances. These results are consistent with known effects of mutation: robustness (maintenance function), evolution or adaptation (emergence of a new feature) and fragility (pathologies). The results also show a weak correlation between the number of amino acid contacts of the mutated amino acid and the amount of structural change induced by its mutation. It is therefore not easy to anticipate the effect of a mutation: The last chapter of my thesis addresses this problem (Chapter 12)
Szilagyi, Ioan. "Technologies sémantiques pour un système actif d’apprentissage." Thesis, Besançon, 2014. http://www.theses.fr/2014BESA1008/document.
Full textLearning methods keep evolving and new paradigms are added to traditional teaching models where the information and communication systems, particularly the Web, are an essential part. In order to improve the processing capacity of information systems, the Semantic Web defines a model for describing resources (Resource Description Framework - RDF), and a language for defining ontologies (Web Ontology Language – OWL). Based on concepts, methods, learning theories, and following a systemic approach, we have used Semantic Web technologies in order to provide a learning system that is able to enrich and personalize the experience of the learner. As a result of our work we are proposing a prototype for an Active Semantic Learning System (SASA). Following the identification and modeling of entities involved in the learning process, we created the following six ontologies that summarize the characteristics of these entities: (1) learner ontology, (2) learning object ontology, (3) learning objective ontology, (4) evaluation object ontology, (5) annotation object ontology and (6) learning framework ontology. Integrating certain rules in the declared ontologies combined with reasoning capacities of the inference engines embedded in the kernel of the SASA, allow the adaptation of learning content to the characteristics of learners. The use of semantic technologies facilitates the identification of existing learning resources on the web as well as the interpretation and aggregation of these resources within the context of SASA
Bar, Romain. "Développement de méthodes d'analyse de données en ligne." Phd thesis, Université de Lorraine, 2013. http://tel.archives-ouvertes.fr/tel-00943148.
Full textBar, Romain. "Développement de méthodes d'analyse de données en ligne." Electronic Thesis or Diss., Université de Lorraine, 2013. http://www.theses.fr/2013LORR0216.
Full textHigh dimensional data are supposed to be independent on-line observations of a random vector. In the second chapter, the latter is denoted by Z and sliced into two random vectors R et S and data are supposed to be identically distributed. A recursive method of sequential estimation of the factors of the projected PCA of R with respect to S is defined. Next, some particular cases are investigated : canonical correlation analysis, canonical discriminant analysis and canonical correspondence analysis ; in each case, several specific methods for the estimation of the factors are proposed. In the third chapter, data are observations of the random vector Zn whose expectation En varies with time. Let Rn = Zn - En be and suppose that the vectors Rn form an independent and identically distributed sample of a random vector R. Stochastic approximation processes are used to estimate on-line direction vectors of the principal axes of a partial principal components analysis (PCA) of ~Z. This is applied next to the particular case of a partial generalized canonical correlation analysis (gCCA) after defining a stochastic approximation process of the Robbins-Monro type to estimate recursively the inverse of a covariance matrix. In the fourth chapter, the case when both expectation and covariance matrix of Zn vary with time n is considered. Finally, simulation results are given in chapter 5
Vervier, Kevin. "Méthodes d’apprentissage structuré pour la microbiologie : spectrométrie de masse et séquençage haut-débit." Thesis, Paris, ENMP, 2015. http://www.theses.fr/2015ENMP0081/document.
Full textUsing high-throughput technologies is changing scientific practices and landscape in microbiology. On one hand, mass spectrometry is already used in clinical microbiology laboratories. On the other hand, the last ten years dramatic progress in sequencing technologies allows cheap and fast characterization of microbial diversity in complex clinical samples. Consequently, the two technologies are approached in future diagnostics solutions. This thesis aims to play a part in new in vitro diagnostics (IVD) systems based on high-throughput technologies, like mass spectrometry or next generation sequencing, and their applications in microbiology.Because of the volume of data generated by these new technologies and the complexity of measured parameters, we develop innovative and versatile statistical learning methods for applications in IVD and microbiology. Statistical learning field is well-suited for tasks relying on high-dimensional raw data that can hardly be used by medical experts, like mass-spectrum classification or affecting a sequencing read to the right organism. Here, we propose to use additional known structures in order to improve quality of the answer. For instance, we convert a sequencing read (raw data) into a vector in a nucleotide composition space and use it as a structuredinput for machine learning approaches. We also add prior information related to the hierarchical structure that organizes the reachable micro-organisms (structured output)
Sokol, Marina. "Méthodes d’apprentissage semi-supervisé basé sur les graphes et détection rapide des nœuds centraux." Thesis, Nice, 2014. http://www.theses.fr/2014NICE4018/document.
Full textSemi-supervised learning methods constitute a category of machine learning methods which use labelled points together with unlabeled data to tune the classifier. The main idea of the semi-supervised methods is based on an assumption that the classification function should change smoothly over a similarity graph. In the first part of the thesis, we propose a generalized optimization approach for the graph-based semi-supervised learning which implies as particular cases the Standard Laplacian, Normalized Laplacian and PageRank based methods. Using random walk theory, we provide insights about the differences among the graph-based semi-supervised learning methods and give recommendations for the choice of the kernel parameters and labelled points. We have illustrated all theoretical results with the help of synthetic and real data. As one example of real data we consider classification of content and users in P2P systems. This application demonstrates that the proposed family of methods scales very well with the volume of data. The second part of the thesis is devoted to quick detection of network central nodes. The algorithms developed in the second part of the thesis can be applied for the selections of quality labelled data but also have other applications in information retrieval. Specifically, we propose random walk based algorithms for quick detection of large degree nodes and nodes with large values of Personalized PageRank. Finally, in the end of the thesis we suggest new centrality measure, which generalizes both the current flow betweenness centrality and PageRank. This new measure is particularly well suited for detection of network vulnerability
Boussetat, Mbaye Sana. "Analyse et critique des méthodes d'une encyclopédie en ligne : Wikipédia." Thesis, Cergy-Pontoise, 2016. http://www.theses.fr/2016CERG0798.
Full textBased on participant observation and collection of data gathered by other observation techniques (statistics, direct observation), we explored in depth the operation of an encyclopedia: the free and open online encyclopedia Wikipedia. The participant observation consists in actively contributing to the open source project, in getting involved in its operation, even to the point of trying to act on some organizational aspects (decision-making, publication of articles), all this in order to analyze the methods used. However, let’s immediately rectify, this is not only an open source project in the strict sense of the term, but a very particular concept of which extent, public, practices, legal nature, make so that it is located at the crossroads of various backgrounds: scientific world, free culture, students, forums’ public and Web 2.0.On the theoretical level, this observation presents several interests. It first allows to show that Wikipedia is not limited to an immaterial computer activity. Wikipedia now penetrates all cultural, scientific and even legal activities.Through this study, we can also learn on the dynamic, the functioning and the credit to grant to such an undertaking. The immersion in the project has indeed helped us to better understand the logics and dynamics of Wikipedia. This study can also serve as reference or comparative element in the observation and analysis of similar projects or in paper form of which construction methods are different (Encyclopædia Universalis, Larousse online encyclopedia, Encarta ...). And finally, in step with the digital age, it offers the prospect of new regulatory pathways and trails to exploit these new cognitive tools
Montalibet, Virginie. "Amélioration du suivi clinique de tumeurs intracrâniennes à l’aide d’équations différentielles et de méthodes d’apprentissage." Electronic Thesis or Diss., Bordeaux, 2024. http://www.theses.fr/2024BORD0174.
Full textMeningiomas are among the most common benign tumors of the central nervous system. Less dangerous than malignant tumors, clinicians generally opt for a regular clinical follow-up rather than treatment or surgery that may be more dangerous than the presence of the meningioma. In this work, in close collaboration with the clinic, we aim to study each stage of patient follow-up to assist surgeons in their decision-making. The first phase, which is necessary to the subsequent ones, involves acquiring MRI images to identify and segment tumors. Here, we consider the application of deep learning algorithms to automate this segmentation, which is currently performed by clinicians. A cohort of 319 patients, along with their images and segmentation masks, will be used for this purpose. An initial analysis will reveal promising results for the identification and segmentation of meningiomas within the cohort. We will compare the architecture used with other models and previous work in the literature. Once the tumor has been segmented and diagnosed, the patient is generally followed up regularly. At each consultation, the doctor takes new images to monitor the tumor’s evolution over time and to build up a cohort of longitudinal data. The clinician must then answer questions such as : should surgery be considered, or should follow-up continue ? A better understanding of meningioma growth would help guide the surgeon’s decision. To this end, we will study and compare different mathematical models based on ordinary differential equations (ODE), able to modeling tumor growth. Various parametric estimation strategies will also be evaluated, primarily in terms of their robustness. In the end, mixed effects and the Gompertz model stand out as providing the best results. Represented by a sigmoid curve and characterized by the decrease of its initial growth rate, the Gompertz model provides additional information. Based on this, and in an attempt to provide patient-specific clinical follow-up, we will try to identify the different phases of this growth, as well as predict the evolution of meningiomas. Mechanistic modeling (ODE) as well as statistical and deep learning methods will be used in this work. Finally, the whole approach will be tested on other tumor types such as schwannomas, or meningiomas induced by hormonal treatments
Vestberg, Francine. "L’interaction comme méthode d’apprentissage du Français langue étrangère (FLE) en Suède." Thesis, Högskolan Dalarna, Franska, 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:du-23450.
Full textMy work is inspired by the idea of the Swedish National Agency for Education (Skolverket, 2011) which suggests that "students should have the opportunity to develop the ability of communication skills and comprehension of the language". My thesis begins with a general introduction, continues with an analysis of the educational materials used in the learning of FLE (Français Langue Etrangère) and activities promoting the interaction – reciprocal exchange which can be verbal or nonverbal, to finally ending with a general conclusion. As there is no teaching without appropriate methods, there are textbooks that can be used to facilitate learning through the interaction of the target language. The activities of oral and written communication of a pedagogic nature are essential. Just like it is written in Revue française de pédagogie (1994:133) "The classroom is a complex social system whose parts are in dynamic interaction, actors (teacher and student[s]), situation, material according to social status." In order for me to do my research, I had to focus my attention on the didactic analysis of the textbooks that I have been enabled to consult. It is two supporting textbooks: Mais oui 3 and Escalade Littéraire. Regarding the analysis of educational materials which encourage to interact, these two textbooks offer exercises of reflection depending on the level of the learner. In my work, I focus on the didactical and the linguistic skills of these two textbooks.
D'Attoma, Amélie. "Développement de méthodes bidimensionnelles en ligne LCxLC-MS pour l'analyse de composés chargés." Phd thesis, Université Claude Bernard - Lyon I, 2013. http://tel.archives-ouvertes.fr/tel-01056279.
Full textOsman, Ousama. "Méthodes de diagnostic en ligne, embarqué et distribué dans les réseaux filaires complexes." Thesis, Université Clermont Auvergne (2017-2020), 2020. http://www.theses.fr/2020CLFAC038.
Full textThe research conducted in this thesis focuses on the diagnosis of complex wired networks using distributed reflectometry. It aims to develop new distributed diagnostic techniques for complex networks that allow data fusion as well as communication between reflectometers to detect, locate and characterize electrical faults (soft and hard faults). This collaboration between reflectometers solves the problem of fault location ambiguity and improves the quality of diagnosis. The first contribution is the development of a graph theory-based method for combining data between distributed reflectometers, thus facilitating the location of the fault. Then, the amplitude of the reflected signal is used to identify the type of fault and estimate its impedance. The latter is based on the regeneration of the signal by compensating for the degradation suffered by the diagnosis signal during its propagation through the network. The second contribution enables data fusion between distributed reflectometers in complex networks affected by multiple faults. To achieve this objective, two methods have been proposed and developed: the first is based on genetic algorithms (GA) and the second is based on neural networks (RN). These tools combined with distributed reflectometryallow automatic detection, location, and characterization of several faults in different types and topologies of wired networks. The third contribution proposes the use of information-carrying diagnosis signal to integrate communication between distributed reflectometers. It properly uses the phases of the MCTDR multi-carrier signal to transmit data. This communication ensures the exchange of useful information (such as fault location and amplitude) between reflectometers on the state of the cables, thus enabling data fusion and unambiguous fault location. Interference problems between the reflectometers are also addressed when they simultaneously inject their test signals into the network. These studies illustrate the efficiency and applicability of the proposed methods. They also demonstrate their potential to improve the performance of the current wired diagnosis systems to meet the need and the problem of detecting and locating faults that manufacturers and users face today in electrical systems to improve their operational safety
D'Attoma, Amélie. "Développement de méthodes bidimensionnelles en ligne LCxLC-MS pour l’analyse de composés chargés." Thesis, Lyon 1, 2013. http://www.theses.fr/2013LYO10214/document.
Full textThis manuscript is dedicated to the development of on-line two-dimensional liquid chromatography for the analysis of charged compounds with mass spectrometry coupling. The context of the study and the theoretical principles of liquid chromatography in one or two dimensions are presented. Experimental conditions such as instrumentation, columns, studied compounds are detailed. Some one-dimensional studies have been experimented to know the kinetic behaviour of small ionisables molecules and peptides depending on conditions. The orthogonality of chromatographic system and generated peak capacity have then been studied in order to know which experimental conditions are the most interesting for the analysis of charged compounds. The comparisons of systems have been done with new descriptor for orthogonality and effective peak capacity. Two-dimensional systems were then set up. Some limits in term of injected volume in the second dimension have been established in RPLC and in HILIC. The usefulness of a split to reduce the injected volume in second dimension has been studied. Some peptides mixtures have been separated by RPLCxRPLC(-MS), RPLCxHILIC(-MS) and HILICxRPLC. The MS coupling has been optimized. The interest of two dimensional separations is underlined compared to classical LC-MS separations
Inghilterra, Xavier. "L’apprenance collective entre pairs à l’aune du modèle transmissif : Impact des dispositifs de partage social sur les communautés d’apprentissage en ligne." Thesis, Toulon, 2016. http://www.theses.fr/2016TOUL0002/document.
Full textThis research is interested in the effects led by the digital devices plans of social sharing on the pratices of collaboration, communication and mediation of students in context of distance learning. The goal is the understand the origin of the collaborative process of collective apprenance which is illustrated in the communities of apprenticeship outside the academic institution. A netnographic observation is conducted whith Bachelor and Master's degree in a private training center ; our corpus is made of 1405 messages taken in the forums of the institutional platform and on Facebook or Google +. We assume the information and communication sociotechnical devices participate in the horizontalisation of student's practices. We highlight the paradox of these learning communities which are, unwittingly, in a process of social domination by having choosing a priori a decentralized structure
Inghilterra, Xavier. "L’apprenance collective entre pairs à l’aune du modèle transmissif : Impact des dispositifs de partage social sur les communautés d’apprentissage en ligne." Electronic Thesis or Diss., Toulon, 2016. http://www.theses.fr/2016TOUL0002.
Full textThis research is interested in the effects led by the digital devices plans of social sharing on the pratices of collaboration, communication and mediation of students in context of distance learning. The goal is the understand the origin of the collaborative process of collective apprenance which is illustrated in the communities of apprenticeship outside the academic institution. A netnographic observation is conducted whith Bachelor and Master's degree in a private training center ; our corpus is made of 1405 messages taken in the forums of the institutional platform and on Facebook or Google +. We assume the information and communication sociotechnical devices participate in the horizontalisation of student's practices. We highlight the paradox of these learning communities which are, unwittingly, in a process of social domination by having choosing a priori a decentralized structure
Carbonnel, Sabine. "Intégration et modélisation de connaissances linguistiques pour la reconnaissance d'écriture manuscrite en-ligne." Rennes, INSA, 2005. http://www.theses.fr/2005ISAR0022.
Full textHandwriting recognition is a difficult problem which cannot be reduced to graphic shapes recognition: it is important to integrate linguistic knowledge to guide the recognition. The objectives of our work are to integrate lexical knowledge to improve a recognition system of on-line handwritten words, taking into account constraints of computing time and memory requirement with the intention of integrate the system on devices with limited capacities. We propose a lexical processing based on a language model of characters n-grams, a modeling to reduce the research space in a lexicon as well as an automatic modeling of an edit distance specific to handwriting. These modelings improve the recognition system on which our work is based, limit the duration of the lexical processing and moreover are easily adaptable to the system evolutions and the context of use
Toure, Adja Adama. "Mise au point de méthodes de mesures rhéologiques en ligne : Application aux mousses liquides." Electronic Thesis or Diss., Université Grenoble Alpes, 2024. http://www.theses.fr/2024GRALI059.
Full textThe application of liquid foams in the agri-food, cosmetics, pharmaceutical and other sectors is becoming increasingly important. Because of their performance, including their low density and texture, they have become interesting materials. Liquid foams continuously formulated in static mixers are complex fluids made up of the dispersion of air bubbles in a solution of surfactants. They are rheologically difficult to characterize at the end of the manufacturing process, due to the various physical phenomena that influence their structural evolution. On-line measurement of their rheological characteristics would make it possible to overcome this difficulty. To achieve this objective, a step-by-step strategy was adopted. First, the hydrodynamics of the static mixers were verified and calibrated. To this end, the static mixers were modeled as a porous medium, due to their geometric composition consisting of a network of intersecting bars forming pore-like channels. Using Newtonian and non-Newtonian viscoelastic fluids, with or without a flow threshold, we acquired data on head loss as a function of flow rate under laminar flow conditions, and measured the characteristic geometric parameters of the static mixer to develop a law for measuring the viscosity of the fluids flowing through it. We proposed a quantitative model based on geometric parameters linking process data (pressure drop and volume flow rate) to rheometric data (first differences in normal stress, stress and shear rate).We have shown that the structure of these foams is driven by shear rheology, which we have characterized independently. Numerical simulation tests with Newtonian fluids of known viscosity supported the systemic rheology results. To support the results of the porous medium approach obtained with the static mixer, we developed a laboratory rotational rheometer at the output of the static mixer, enabling us to measure the stress-shear relationship from torque and rotational speed data of foam samples flowing between two coaxial cylinders. The results obtained did not allow us to estimate the viscosity of the foams, due to the characteristic limits of the device used, but also to the effect of the flow rate of foam arriving in the rotational device.Key words: Liquid foams. Rheology, viscoelasticity. Static mixers. Porous media, Couette flow
Mescco, Amilcar. "Etude des émissions électromagnétiques CPL large-bande : caractérisation, modélisation et méthodes de mitigation." Télécom Bretagne, 2013. http://www.telecom-bretagne.eu/publications/publication.php?idpublication=14211.
Full textWith the growing demand for broadband applications and reliable connections for command and control systems, PLC technology is playing an increasingly important role in innovative systems such as communication networks and smart grids. The main advantage of this technology is its ability to benefit from the infrastructure of the existing power grid for transmission of electromagnetic signals. Currently, broadband systems mainly operate in the frequency range from 2 MHz to 30 MHz. However, in indoor environment of home or office the power supply electrical wires were not originally designed to convey communication signals at high frequencies. This thesis focuses on one of the main limitations to the indoor PLC technology, namely, unintentional radiation of electromagnetic signals. The first objective is to characterize the radiation level of a typical PLC system to check if it allows coexistence with other systems in the same environment. Then, the thesis evaluates and proposes methods to reduce the level of unwanted radiation
Ben, Soussia Amal. "Analyse prédictive des données d’apprentissage, en situation d’enseignement à distance." Electronic Thesis or Diss., Université de Lorraine, 2022. http://www.theses.fr/2022LORR0216.
Full textOver the past few decades, the adoption of e-learning has evolved rapidly and its use has been pushedeven further with the COVID-19 pandemic. The objective of this learning mode is to guarantee thecontinuity of the learning process. However, the online learning is facing several challenges, and themost widespread is the high failure rates among learners. This issue is due to many reasons such asthe heterogeneity of the learners and the diversity of their learning behaviors, their total autonomy, thelack and/or the inefficiency of the pedagogical provided follow-up. . .. Therefore, teachers need a systembased on analytical and intelligent methods allowing them an accurate and early prediction of at-risk offailure learners. This solution is commonly adopted in the state of the art. However, the work carried outdoes not respond to some particularities of the learning process (the continuity and evolution of learning,the diversity of learners and their total autonomy) and to some teachers expectations such as the alertgeneration.This thesis belongs to the field of learning analytics and uses the numeric traces of online learnersto design a predictive system (Early Warning Systems (EWS)) dedicated to teachers in online establish-ments. The objective of this EWS is to identify learners at risk as soon as possible in order to alertteachers about them. In order to achieve this objective, we have dealt with several sub-problems whichhave allowed us to elaborate four scientific contributions.We start by proposing an in-depth methodology based on the Machine Learning (ML) steps and thatallows the identification of four learning indicators among : performance, engagement, reactivity andregularity. This methodology also highlights the importance of temporal data for improving predictionperformance. In addition, this methodology allowed to define the model with the best ability to identifyat-risk learners.The 2nd contribution consists in proposing a temporal evaluation of the EWS using temporal metricswhich measure the precocity of the predictions and the stability of the system. From these two metrics,we study the trade-offs that exist between ML precision metrics and temporal metrics.Online learners are characterized by the diversity of their learning behaviors. Thus, an EWS shouldrespond to this diversity by ensuring an equitable functioning with the different learners profiles. Wepropose an evaluation methodology based on the identification of learner profiles and that uses a widespectrum of temporal and precision metrics.By using an EWS, teachers expect an alert generation. For this reason, we design an algorithm which,based on the results of the prediction, the temporal metrics and the notion of alert rules, proposes anautomatic method for alert generation. This algorithm targets mainly at-risk learners.The context of this thesis is the French National Center for Distance Education (CNED). In parti-cular, we use the numeric traces of k-12 learners enrolled during the 2017-2018 and 2018-2019 schoolyears
Ghourabi, Samira. "L’appropriation des référentiels normatifs dans les environnements d’apprentissage en ligne : evaluation de l’usage de la plate forme Moodle de l’Université Virtuelle de Tunis." Thesis, Lille 3, 2016. http://www.theses.fr/2016LIL30003.
Full textWith the significant growth of Information and Communication Technologies, the traditional methods of disseminating, accessing and sharing information have been reshaped.Currently, the ICT dedicated to education are thriving, enabling learner human new ways to train independently and draw their own learning paths. The e-Learning is a new form of learning facilitates both, access to educational resources, sharing and remote collaboration.To overcome the difficulties related to interoperability of e-Learning systems and guarantee the quality of learning, the application of normative references must apply to all aspects of a learning device (learner’s profiles, methods of monitoring and evaluation, communication tools and capabilities ...). The work proposed in this thesis aims to examine the state of reference standards application in the Virtual University of Tunis. Through a case study, we examined on the one hand, the receptivity of the VUT on regarding the interests of normative references dedicated to education, on the other hand; we identified the communication tools integrated in the platform Moodle allowing the tutor to follow up, evaluation and support
Passard, Christian. "Application des méthodes d'interrogation neutronique active à l'analyse en ligne dans les usines de retraitement." Grenoble 1, 1993. http://www.theses.fr/1993GRE10034.
Full textPakdel, Ali. "De l'activité communicative à l'activité sociale d'apprentissage des langues en ligne : analyse de la dynamique sociale en contexte institutionnel." Phd thesis, Université de Provence - Aix-Marseille I, 2011. http://tel.archives-ouvertes.fr/tel-00637137.
Full textDelest, Sébastien. "Segmentation de maillages 3D à l'aide de méthodes basées sur la ligne de partage des eaux." Phd thesis, Université François Rabelais - Tours, 2007. http://tel.archives-ouvertes.fr/tel-00211378.
Full textNous proposons dans un premier temps une étude assez large des méthodes de segmentation de maillages polygonaux. Nous abordons les algorithmes pour les deux principales familles de méthodes que sont la segmentation en carreaux surfaciques et la segmentation en parties significatives. Nous avons concentré nos travaux sur la ligne de partage des eaux (LPE) et formulé des propositions originales pour la fonction de hauteur de la LPE et des stratégies pour limiter la sur-segmentation que produit naturellement la LPE.
Sarrut, Morgan. "Optimisation de méthodes bidimensionnelles en ligne LCxLC-UV/MS et LCxSFC-UV pour l’analyse d’échantillons complexes." Thesis, Lyon, 2016. http://www.theses.fr/2016LYSE1189/document.
Full textComprehensive two-dimensional liquid chromatography is a powerful but complex separative technique. After detailing the interest of such a technique, the method development issues and the experimental conditions employed throughout this work, a particular attention is paid to the optimization of LCxLC methods. Accordingly an optimization procedure based on Pareto-optimal method is described. The predicted optimal conditions are then applied to experimental RPLCxRPLC separations of complex samples of peptides and compared with 1D-RPLC in terms of peak capacity, analysis time and sensitivity clearly showing the advantage of RPLCxRPLC approach.The optimization of a HICxRPLC-UV/MS method for the exhaustive characterization of an antibody-drug conjugate is achieved highlighting the high complementarity of the different detection modes used both in 1D and 2D. Finally, a proof of concept concerning the implementation of RPLCxSFC coupling is achieved with the aim of increasing the separation space coverage for neutral compounds. The optimized RPLCxSFC separation is then compared with an optimized RPLCxRPLC approach for the analysis of a bio-oil sample showing that RPLCxSFC is a credible alternative for the separation of such a sample
Delest, Sébastien. "Ségmentation de maillages 3D à l'aide de méthodes basées sur la ligne de partage des eaux." Tours, 2007. http://www.theses.fr/2007TOUR4025.
Full textMesh segmentation is a necessary tool for many applications. The mesh is decomposed into several regions from surface or shape information. In the last several years, many algorithms have been proposed in this growing area, with applications in many different areas as 3D shape matching and retrieval, compression, metamorphosis, collision detection, texture mapping, simplification, etc. First, we propose a review of mesh segmentation methods. We discuss about the algorithms relative to the two main types of methods: the patch-type segmentation and the part-type segmentation. We focused on the watershed transformation and proposed new approches relativing to the height function and strategies to avoid over segmentation produced by the watershed
Zeaiter, Magida. "Mesures robustes en ligne des solutés organiques par spectrométrie infrarouge et étalonnages multivariés." Montpellier 2, 2004. http://www.theses.fr/2004MON20186.
Full textMichel, Thierry. "Test en ligne des systèmes à base de microprocesseur." Phd thesis, Grenoble INPG, 1993. http://tel.archives-ouvertes.fr/tel-00343488.
Full textAmor, Yasmine. "Ιntelligent apprοach fοr trafic cοngestiοn predictiοn." Electronic Thesis or Diss., Normandie, 2024. http://www.theses.fr/2024NORMR129.
Full textTraffic congestion presents a critical challenge to urban areas, as the volume of vehicles continues to grow faster than the system’s overall capacity. This growth impacts economic activity, environmental sustainability, and overall quality of life. Although strategies for mitigating traffic congestion have seen improvements over the past few decades, many cities still struggle to manage it effectively. While various models have been developed to tackle this issue, existing approaches often fall short in providing real-time, localized predictions that can adapt to complex and dynamic traffic conditions. Most rely on fixed prediction horizons and lack the intelligent infrastructure needed for flexibility. This thesis addresses these gaps by proposing an intelligent, decentralized, infrastructure-based approach for traffic congestion estimation and prediction.We start by studying Traffic Estimation. We examine the possible congestion measures and data sources required for different contexts that may be studied. We establish a three-dimensional relationship between these axes. A rule-based system is developed to assist researchers and traffic operators in recommending the most appropriate congestion measures based on the specific context under study. We then proceed to Traffic Prediction, introducing our DECentralized COngestion esTimation and pRediction model using Intelligent Variable Message Signs (DECOTRIVMS). This infrastructure-based model employs intelligent Variable Message Signs (VMSs) to collect real-time traffic data and provide short-term congestion predictions with variable prediction horizons.We use Graph Attention Networks (GATs) due to their ability to capture complex relationships and handle graph-structured data. They are well-suited for modeling interactions between different road segments. In addition to GATs, we employ online learning methods, specifically, Stochastic Gradient Descent (SGD) and ADAptive GRAdient Descent (ADAGRAD). While these methods have been successfully used in various other domains, their application in traffic congestion prediction remains under-explored. In our thesis, we aim to bridge that gap by exploring their effectiveness within the context of real-time traffic congestion forecasting.Finally, we validate our model’s effectiveness through two case studies conducted in Muscat, Oman, and Rouen, France. A comprehensive comparative analysis is performed, evaluating various prediction techniques, including GATs, Graph Convolutional Networks (GCNs), SGD and ADAGRAD. The achieved results underscore the potential of DECOTRIVMS, demonstrating its potential for accurate and effective traffic congestion prediction across diverse urban contexts
Dabbebi, Ines. "Conception et génération dynamique de tableaux de bord d’apprentissage contextuels." Thesis, Le Mans, 2019. http://www.theses.fr/2019LEMA1040/document.
Full textThis work is part of a broader issue of Learning Analytics (LA). It is particularly carried out within the context of the HUBBLE project, a national observatory for the design and sharing of data analysis processes. We are interested in communicating data analysis results to users by providing LA dashboards (LAD). Our main issue is the identification of generic LAD structures in order to generate dynamically tailored LAD. These structures must be generic to ensure their reuse, and adaptable to users’ needs. Existing works proposed LAD which remains too general or developed in an adhoc way. According to the HUBBLE project, we want to use identified decisions of end-users to generate dynamically our LAD. We were interested in the business intelligence area because of the place of dashboards in the decision-making process. Decision-making requires an explicit understanding of user needs. That's why we have adopted a user-centered design (UCD) approach to generate adapted LAD. We propose a new process for capturing end-users’ needs, in order to elaborate some models (Indicator, visualization means, user, pattern, …). These models are used by a generation process implemented in a LAD dynamic generator prototype. We conducted an iterative evaluation phase. The objective is to refine our models and validate the efficiency of our generation process. The second iteration demonstrates the impact of the decision on the LAD generation. Thus, we can confirm that the decision is considered as a central element for the generation of LADs
Iguiniz, Marion. "Développement de méthodes bidimensionnelles en ligne LCxLC-UV/MS et LCxSFC-UV pour l’analyse de composés pharmaceutiques." Thesis, Lyon, 2018. http://www.theses.fr/2018LYSE1200/document.
Full textTwo-dimensional liquid chromatography (2D-LC) is a powerful technique considering its high separation power. After showing the advantage of 2D-LC in the pharmaceutical area and presenting the challenges related to quantitative analysis, special attention was paid to method development. With the aim of developing a generic analytical strategy for pharmaceuticals, the first step of our approach consisted in selecting a set of three 2D-systems with the help of a methodology previously developed. In a second step, the potential of these 2D-systems was evaluated for the purpose of quantitative analysis. An analytical strategy able to be applied to pharmaceutical analysis in an industrial context was proposed. Finally, the potential of RPLCxSFC was investigated in two different cases. Firstly, for comparing this on-line two dimensional technique to on-line RPLCxRPLC with respect of the separation power. Secondly, for chiral compounds by developing a selective RPLCxSFC method for simultaneous achiral-chiral analysis. The advantage of such method was highlighted by comparing to conventional approaches
Renau-Ferrer, Ney. "Outils et méthodes pour l'analyse automatique multi-niveaux de tracés manuscrits à caractère géométrique acquis en ligne." Thesis, Antilles-Guyane, 2011. http://www.theses.fr/2010AGUY0394/document.
Full textThis thesis handles the problem of the automatic analysis of online hand drawn geometric sketches. An online sketch can be analysed according to several points of view. As for offline sketching, we can try to recognize the produced shape. However, online sketching allows other levels of analysis. For example the analysis of the behavior of the drawer during the production of the sketch. In this thesis, we have tried to develop tools allowing a multi level analysis, including both shape and behavior analysis. The first part of our work deals with the pre treatments that must be performed on the sketch in order to allow upper level analysis. Those pre treatments are filtering, mixed segmentation and feature points detection and labelisation. In the second part, we approach shape analysis in two aspects: shape recognition and evaluation. We have developed a appearance based method which use local descriptors to allows both recognition and evaluation of the quality of a produced shape compared to the model . in the last part we propose a method for drawer's behavior extraction and modeling. Then we show how we can not only determine the favorite procedure of a drawer but also recognize the drawer by analyzing his behavior
Picque, Daniel. "Développement de méthodes de mesure en ligne de la concentration en éthanol dans les procédés de fermentation." Compiègne, 1989. http://www.theses.fr/1989COMPD286.
Full textLaubie, Raphaëlle. "Les Déterminants de l’Action Collective en Ligne dans les Communautés Virtuelles de Patients : une Approche Multi-Méthodes." Thesis, Paris Sciences et Lettres (ComUE), 2017. http://www.theses.fr/2017PSLED036/document.
Full textOver the last few years, virtual patients’communities have been developing tremendously over the Internet. These Web 2.0 communities allow frequent interactions among patients, who can share health-related information within an interactive environment. While many agree on the opportunity represented by those communities for its users, we know very little about what determines patients’ online collective action, specifically on virtual communities as well as the fundamentals of online collective action in these virtual spaces. Accordingly, this doctoral work examines why patients interact with others and how they interact on topics related to their disease through these virtual communities. Drawing on the goal-directed behavior (MGB), the expectancy-value (EVT) theories, the field force theory, gift concepts and field interviews, we have developed a model for examining patients’ online interactions and identified gift-giving behaviors in the context of online collective action. A multi-method, qualitative and quantitative approaches, enables us to explore patients’ interactions and measures the determinants of online collective action on these virtual spaces. The qualitative analysis of 54 interviews conducted with patients, patient’s relatives, Health 2.0 professionals, doctors and caregivers allows refining the research model, which has then been tested through a survey handled with 269 patients, members of patient’s communities. This research contributes to IS research by increasing our knowledge regarding the individual dynamics and interactions that surround online patients’ communities
Bérujon, Sébastien. "Métrologie en ligne de faisceaux et d'optiques X de synchrotrons." Phd thesis, Université de Grenoble, 2013. http://tel.archives-ouvertes.fr/tel-00859120.
Full textAtine, Jean-Charles. "Méthodes d'apprentissage flou : application à la segmentation d'images biologiques." Toulouse, INSA, 2005. http://eprint.insa-toulouse.fr/archive/00000272/.
Full textThe presented works have for objective to help the biologists in the diagnosis of the cellular viability by using some methods of classification. Our work announces a strategy of classification allowing to building partition of images of cells coming from an optical microscope. We classify automatically the cells by operating the segmentation on images using the developed algorithm T-LAMDA. A statement concerning the existing classification methods, the color space and the resistance to noise, allows to finding the structure the most adapted to our study. The comparative analysis of various methods (of which LAMDA and T-LAMDA methods), allows us to put in evidence the most appropriate for the classification of cells subjected to the blue of methylene solution. We propose some supervised algorithms based on LAMDA to show if the way of treating the data influence the result. The T-LAMDA algorithm, based on the decision trees, shows itself the best adapted for our study and so gives more precise results than other methods, with a shorter time of execution. We suggest learning by using the CELCA application, Cell Classification Application, which uses the developed T-LAMDA algorithm. The software takes care of calculations of the kinetics, according to the images which respect to a well defined protocol. Time for treating 117 images is 6 '47'' minutes, what is widely below the time taken by biologists to count the cells
Belmonte, Mylène. "Automatisation intégrale de la ligne 1 : étude et modélisation du trafic mixte." Compiègne, 2008. http://www.theses.fr/2008COMP1752.
Full textThe current thesis takes place in the framework of fine 1 automation project. It is a project that consists of upgrading fine 1 (with drivers on-board) to unattended train operation (no drivers or agents on-board) without service interruption. It results a mixed fleet operation for a period of two years between manually driven trains and driverless ones. In this context, my research work consists of continuing in a first step the development of software destined to study and analyse the mixed fleet operation on fine 1. Once operational; the model has served as a simulation in two case studies that I elaborate in a second step. These two case studies aim to answer problematical matters affecting the management of the mixed trains running inside the terminal stations. They take into account two opposite constraints: people-staff safety on one hand and quality of service offered to passengers on the other hand. The analysis and the results of these two case studies will be described in the current PhD report
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.
Full textIn 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
Farhat, Ramzi. "Approche d'assistance aux auteurs pour la réutilisation d'objets d'apprentissage." Phd thesis, Institut National des Télécommunications, 2010. http://tel.archives-ouvertes.fr/tel-00589601.
Full textEl, Sahmarany Lola. "Méthodes d'amélioration pour le diagnostic de câble par réflectométrie." Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2013. http://tel.archives-ouvertes.fr/tel-00999462.
Full textMazet, Vincent. "Développement de méthodes de traitement de signaux spectroscopiques : estimation de la ligne de base et du spectre de raies." Phd thesis, Université Henri Poincaré - Nancy I, 2005. http://tel.archives-ouvertes.fr/tel-00011477.
Full textDans un premier temps est proposée une méthode déterministe qui permet d'estimer la ligne de base des spectres par le polynôme qui minimise une fonction-coût non quadratique (fonction de Huber ou parabole tronquée). En particulier, les versions asymétriques sont particulièrement bien adaptées pour les spectres dont les raies sont positives. Pour la minimisation, on utilise l'algorithme de minimisation semi-quadratique LEGEND.
Dans un deuxième temps, on souhaite estimer le spectre de raies : l'approche bayésienne couplée aux techniques MCMC fournit un cadre d'étude très efficace. Une première approche formalise le problème en tant que déconvolution impulsionnelle myope non supervisée. En particulier, le signal impulsionnel est modélisé par un processus Bernoulli-gaussien à support positif ; un algorithme d'acceptation-rejet mixte permet la simulation de lois normales tronquées. Une alternative intéressante à cette approche est de considérer le problème comme une décomposition en motifs élémentaires. Un modèle original est alors introduit ; il a l'intérêt de conserver l'ordre du système fixe. Le problème de permutation d'indices est également étudié et un algorithme de ré-indexage est proposé.
Les algorithmes sont validés sur des spectres simulés puis sur des spectres infrarouge et Raman réels.
Hirwa, Serge. "Méthodes de commande avancées appliquées aux viseurs." Phd thesis, Supélec, 2013. http://tel.archives-ouvertes.fr/tel-00969110.
Full textAlattar, Farah Basma. "Débats participatifs en ligne et hors ligne en milieu scolaire - Pour une éducation à la citoyenneté." Thesis, Paris 3, 2020. http://www.theses.fr/2020PA030001.
Full textOur action-research refers to active and project pedagogy. We have considered ways of developing participatory debates in schools, using a digital environment; we have based this reflection on a concrete experience conducted from a Parisian high school: a multilingual simulation of international parliaments that we have named "Parliament of Future Citizens".We first highlighted the role of these discussions and consultations in raising students' awareness of the need to build a more humane world. We have complemented this training in debates with the creation of a Non-Governmental Organization in the field of health. The objective was to give citizenship education its deep meaning by allowing Spanish, French, Italian, Romanian and Turkish pupils to move from a virtual space of discussion to an associative space of mediation that works for solidarity actions. Finally, through the establishment of a Hackathon with European students, we have demonstrated that competition and democratic debates aimed at inclusive education are complementary and mutually developing.Managing an international educational environment for online and offline debates in high-school requires the teacher to implement rigorous social engineering, training and education, thought in terms of tools, planning and human resources management. This hybrid education organization must be able to combine online devices with dedicated tools ranging from content aggregators, forums, training sites, online voting to the organization of the platform dedicated to e-learning
Langlet, Alyssa. "Développement de méthodes d'analyse de comprimés à haute vitesse." Mémoire, Université de Sherbrooke, 2018. http://hdl.handle.net/11143/11773.
Full text