Dissertationen zum Thema „Modèle bayésien non paramétrique“
Geben Sie eine Quelle nach APA, MLA, Chicago, Harvard und anderen Zitierweisen an
Machen Sie sich mit Top-50 Dissertationen für die Forschung zum Thema "Modèle bayésien non paramétrique" bekannt.
Neben jedem Werk im Literaturverzeichnis ist die Option "Zur Bibliographie hinzufügen" verfügbar. Nutzen Sie sie, wird Ihre bibliographische Angabe des gewählten Werkes nach der nötigen Zitierweise (APA, MLA, Harvard, Chicago, Vancouver usw.) automatisch gestaltet.
Sie können auch den vollen Text der wissenschaftlichen Publikation im PDF-Format herunterladen und eine Online-Annotation der Arbeit lesen, wenn die relevanten Parameter in den Metadaten verfügbar sind.
Sehen Sie die Dissertationen für verschiedene Spezialgebieten durch und erstellen Sie Ihre Bibliographie auf korrekte Weise.
Rivoirard, Vincent. „Estimation bayésienne non paramétrique“. Phd thesis, Université Paris-Diderot - Paris VII, 2002. http://tel.archives-ouvertes.fr/tel-00002149.
Der volle Inhalt der QuelleSodjo, Jessica. „Modèle bayésien non paramétrique pour la segmentation jointe d'un ensemble d'images avec des classes partagées“. Thesis, Bordeaux, 2018. http://www.theses.fr/2018BORD0152/document.
Der volle Inhalt der QuelleThis work concerns the joint segmentation of a set images in a Bayesian framework. The proposed model combines the hierarchical Dirichlet process (HDP) and the Potts random field. Hence, for a set of images, each is divided into homogeneous regions and similar regions between images are grouped into classes. On the one hand, thanks to the HDP, it is not necessary to define a priori the number of regions per image and the number of classes, common or not.On the other hand, the Potts field ensures a spatial consistency. The arising a priori and a posteriori distributions are complex and makes it impossible to compute analytically estimators. A Gibbs algorithm is then proposed to generate samples of the distribution a posteriori. Moreover,a generalized Swendsen-Wang algorithm is developed for a better exploration of the a posteriori distribution. Finally, a sequential Monte Carlo sampler is defined for the estimation of the hyperparameters of the model.These methods have been evaluated on toy examples and natural images. The choice of the best partition is done by minimization of a numbering free criterion. The performance are assessed by metrics well-known in statistics but unused in image segmentation
Elvira, Clément. „Modèles bayésiens pour l’identification de représentations antiparcimonieuses et l’analyse en composantes principales bayésienne non paramétrique“. Thesis, Ecole centrale de Lille, 2017. http://www.theses.fr/2017ECLI0016/document.
Der volle Inhalt der QuelleThis thesis proposes Bayesian parametric and nonparametric models for signal representation. The first model infers a higher dimensional representation of a signal for sake of robustness by enforcing the information to be spread uniformly. These so called anti-sparse representations are obtained by solving a linear inverse problem with an infinite-norm penalty. We propose in this thesis a Bayesian formulation of anti-sparse coding involving a new probability distribution, referred to as the democratic prior. A Gibbs and two proximal samplers are proposed to approximate Bayesian estimators. The algorithm is called BAC-1. Simulations on synthetic data illustrate the performances of the two proposed samplers and the results are compared with state-of-the art methods. The second model identifies a lower dimensional representation of a signal for modelisation and model selection. Principal component analysis is very popular to perform dimension reduction. The selection of the number of significant components is essential but often based on some practical heuristics depending on the application. Few works have proposed a probabilistic approach to infer the number of significant components. We propose a Bayesian nonparametric principal component analysis called BNP-PCA. The proposed model involves an Indian buffet process to promote a parsimonious use of principal components, which is assigned a prior distribution defined on the manifold of orthonormal basis. Inference is done using MCMC methods. The estimators of the latent dimension are theoretically and empirically studied. The relevance of the approach is assessed on two applications
Autin, Florent. „Point de vue maxiset en estimation non paramétrique“. Phd thesis, Université Paris-Diderot - Paris VII, 2004. http://tel.archives-ouvertes.fr/tel-00008542.
Der volle Inhalt der QuelleNaulet, Zacharie. „Développement d'un modèle particulaire pour la régression indirecte non paramétrique“. Thesis, Paris Sciences et Lettres (ComUE), 2016. http://www.theses.fr/2016PSLED057/document.
Der volle Inhalt der QuelleThis dissertation deals with Bayesian nonparametric statistics, in particular nonparametric mixture models. The manuscript is divided into a general introduction and three parts on rather different aspects of mixtures approaches (sampling, asymptotic, inverse problem). In mixture models, the parameter to infer from the data is a function. We set a prior distribution on an abstract space of functions through a stochastic integral of a kernel with respect to a random measure. Usually, mixture models were used primilary in probability density function estimation problems. One of the contributions of the present manuscript is to use them in regression problems.In this context, we are essentially concerned with the following problems :- Sampling of the posterior distribution- Asymptotic properties of the posterior distribution- Inverse problems, in particular the estimation of the Wigner distribution from Quantum Homodyne Tomography measurements
Gayraud, Ghislaine. „Vitesses et procédures statistiques minimax dans des problèmes d'estimation et des tests d'hypothèses“. Habilitation à diriger des recherches, Université de Rouen, 2007. http://tel.archives-ouvertes.fr/tel-00207687.
Der volle Inhalt der QuelleLa première thèmatique porte sur la résolution via l'approche minimax de divers problèmes d'estimation et de tests d'hypothèses dans un cadre non-paramétrique.
En statistique Bayésienne non-paramétrique, je me suis intéressée à un problème d'estimation d'ensembles à niveau. Les résultats obtenus résultent de l'étude des propriétés asymptotiques d'estimation Bayésienne d'ensembles à niveau. Ce sont des résultats généraux au sens où la consistance et la vitesse de convergence de l'estimateur Bayésien sont établies pour une large classe de lois a priori.
La troisième thématique concerne un problème d'estimation paramétrique dans un modèle de déconvolution aveugle bruitée : il s'agit de restituer la loi du signal entrant. La consistance ainsi que la distribution asymptotique d'une nouvelle procédure d'estimation sont établies.
Okabe, Shu. „Modèles faiblement supervisés pour la documentation automatique des langues“. Electronic Thesis or Diss., université Paris-Saclay, 2023. http://www.theses.fr/2023UPASG091.
Der volle Inhalt der QuelleIn the wake of the threat of extinction of half of the languages spoken today by the end of the century, language documentation is a field of linguistics notably dedicated to the recording, annotation, and archiving of data. In this context, computational language documentation aims to devise tools for linguists to ease several documentation steps through natural language processing approaches.As part of the CLD2025 computational language documentation project, this thesis focuses mainly on two tasks: word segmentation to identify word boundaries in an unsegmented transcription of a recorded sentence and automatic interlinear glossing to predict linguistic annotations for each sentence unit.For the first task, we improve the performance of the Bayesian non-parametric models used until now through weak supervision. For this purpose, we leverage realistically available resources during documentation, such as already-segmented sentences or dictionaries. Since we still observe an over-segmenting tendency in our models, we introduce a second segmentation level: the morphemes. Our experiments with various types of two-level segmentation models indicate a slight improvement in the segmentation quality. However, we also face limitations in differentiating words from morphemes, using statistical cues only. The second task concerns the generation of either grammatical or lexical glosses. As the latter cannot be predicted using training data solely, our statistical sequence-labelling model adapts the set of possible labels for each sentence and provides a competitive alternative to the most recent neural models
Vernet, Elodie Edith. „Modèles de mélange et de Markov caché non-paramétriques : propriétés asymptotiques de la loi a posteriori et efficacité“. Thesis, Université Paris-Saclay (ComUE), 2016. http://www.theses.fr/2016SACLS418/document.
Der volle Inhalt der QuelleLatent models have been widely used in diverse fields such as speech recognition, genomics, econometrics. Because parametric modeling of emission distributions, that is the distributions of an observation given the latent state, may lead to poor results in practice, in particular for clustering purposes, recent interest in using non parametric latent models appeared in applications. Yet little thoughts have been given to theory in this framework. During my PhD I have been interested in the asymptotic behaviour of estimators (in the frequentist case) and the posterior distribution (in the Bayesian case) in two particuliar non parametric latent models: hidden Markov models and mixture models. I have first studied the concentration of the posterior distribution in non parametric hidden Markov models. More precisely, I have considered posterior consistency and posterior concentration rates. Finally, I have been interested in efficient estimation of the mixture parameter in semi parametric mixture models
Mismer, Romain. „Convergence et spike and Slab Bayesian posterior distributions in some high dimensional models“. Thesis, Sorbonne Paris Cité, 2019. http://www.theses.fr/2019USPCC064.
Der volle Inhalt der QuelleThe first main focus is the sparse Gaussian sequence model. An Empirical Bayes approach is used on the Spike and Slab prior to derive minimax convergence of the posterior second moment for Cauchy Slabs and a suboptimality result for the Laplace Slab is proved. Next, with a special choice of Slab convergence with the sharp minimax constant is derived. The second main focus is the density estimation model using a special Polya tree prior where the variables in the tree construction follow a Spike and Slab type distribution. Adaptive minimax convergence in the supremum norm of the posterior distribution as well as a nonparametric Bernstein-von Mises theorem are obtained
Li, Shuxian. „Modélisation spatio-temporelle pour l'esca de la vigne à l'échelle de la parcelle“. Thesis, Bordeaux, 2015. http://www.theses.fr/2015BORD0313/document.
Der volle Inhalt der QuelleEsca grapevine disease is one of the incurable dieback disease with the etiology not completely elucidated. It represents one of the major threats for viticulture around the world. To better understand the underlying process of esca spread and the risk factors of this disease, we carried out quantitative analyses of the spatio-temporal development of esca at vineyard scale. In order to detect the spatial correlation among the diseased vines, the non-parametric statistical tests were applied to the spatio-temporal data of esca foliar symptom expression for 15 vineyards in Bordeaux region. Among vineyards, a large range of spatial patterns, from random to strongly structured, were found. In the vineyards with strongly aggregated patterns, no significant increase in the size of cluster and local spread from symptomatic vines was shown, suggesting an effect of the environment in the explanation of this aggregation. To model the foliar symptom occurrence, we developed hierarchical logistic regression models by integrating exogenous covariates, covariates of neighboring symptomatic vines already diseased, and also a latent process with spatio-temporal auto-correlation. The Bayesian inferences of these models were performed by INLA (Inverse Nested Laplace Approximation) approach. The results confirmed the effect of environmental factors on the occurrence risk of esca symptom. The secondary locally spread of esca from symptomatic vines located on the same row or out of row was not shown. A two-step centered auto-logistic regression model, which explicitly integrated the spatio-temporal neighboring structure, was also developed. At last, a geostatistical method was proposed to interpolate data with a particular anisotropic structure. It allowed interpolating the ancillary variable, electrical resistivity of soil, which were used to estimate the available soil water content at vine-scale. These geostatistical methods and spatio-temporal statistical methods developed in this thesis offered outlook to identify risk factors, and thereafter to predict the development of esca grapevine disease in different agronomical contexts
Löser, Kevin. „Apprentissage non-supervisé de la morphologie des langues à l’aide de modèles bayésiens non-paramétriques“. Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLS203/document.
Der volle Inhalt der QuelleA crucial issue in statistical natural language processing is the issue of sparsity, namely the fact that in a given learning corpus, most linguistic events have low occurrence frequencies, and that an infinite number of structures allowed by a language will not be observed in the corpus. Neural models have already contributed to solving this issue by inferring continuous word representations. These continuous representations allow to structure the lexicon by inducing semantic or syntactic similarity between words. However, current neural models only partially solve the sparsity issue, due to the fact that they require a vectorial representation for every word in the lexicon, but are unable to infer sensible representations for unseen words. This issue is especially present in morphologically rich languages, where word formation processes yield a proliferation of possible word forms, and little overlap between the lexicon observed during model training, and the lexicon encountered during its use. Today, several languages are used on the Web besides English, and engineering translation systems that can handle morphologies that are very different from western European languages has become a major stake. The goal of this thesis is to develop new statistical models that are able to infer in an unsupervised fashion the word formation processes underlying an observed lexicon, in order to produce morphological analyses of new unseen word forms
Hadrich, Ben Arab Atizez. „Étude des fonctions B-splines pour la fusion d'images segmentées par approche bayésienne“. Thesis, Littoral, 2015. http://www.theses.fr/2015DUNK0385/document.
Der volle Inhalt der QuelleIn this thesis we are treated the problem of nonparametric estimation probability distributions. At first, we assumed that the unknown density f was approximated by a basic mixture quadratic B-spline. Then, we proposed a new estimate of the unknown density function f based on quadratic B-splines, with two methods estimation. The first is based on the maximum likelihood method and the second is based on the Bayesian MAP estimation method. Then we have generalized our estimation study as part of the mixture and we have proposed a new estimator mixture of unknown distributions based on the adapted estimation of two methods. In a second time, we treated the problem of semi supervised statistical segmentation of images based on the hidden Markov model and the B-sline functions. We have shown the contribution of hybridization of the hidden Markov model and B-spline functions in unsupervised Bayesian statistical image segmentation. Thirdly, we presented a fusion approach based on the maximum likelihood method, through the nonparametric estimation of probabilities, for each pixel of the image. We then applied this approach to multi-spectral and multi-temporal images segmented by our nonparametric and unsupervised algorithm
Du, Rocher Martin. „Méthode de Denton et modèle non-paramétrique d'étalonnage“. Mémoire, Université de Sherbrooke, 2009. http://savoirs.usherbrooke.ca/handle/11143/4831.
Der volle Inhalt der QuelleRaoux, Jean-Jacques. „Modélisation non-linéaire des composants électroniques : du modèle analytique au modèle tabulaire paramétrique“. Limoges, 1995. http://www.theses.fr/1995LIMO0006.
Der volle Inhalt der QuelleMohdeb, Zaher. „Tests d'hypothèses linéaires dans un modèle de régression non paramétrique“. Versailles-St Quentin en Yvelines, 1999. http://www.theses.fr/1999VERS0003.
Der volle Inhalt der QuelleArbel, Julyan. „Contributions à la statistique bayésienne non-paramétrique“. Phd thesis, Université Paris Dauphine - Paris IX, 2013. http://tel.archives-ouvertes.fr/tel-01067718.
Der volle Inhalt der QuelleBartcus, Marius. „Bayesian non-parametric parsimonious mixtures for model-based clustering“. Thesis, Toulon, 2015. http://www.theses.fr/2015TOUL0010/document.
Der volle Inhalt der QuelleThis thesis focuses on statistical learning and multi-dimensional data analysis. It particularly focuses on unsupervised learning of generative models for model-based clustering. We study the Gaussians mixture models, in the context of maximum likelihood estimation via the EM algorithm, as well as in the Bayesian estimation context by maximum a posteriori via Markov Chain Monte Carlo (MCMC) sampling techniques. We mainly consider the parsimonious mixture models which are based on a spectral decomposition of the covariance matrix and provide a flexible framework particularly for the analysis of high-dimensional data. Then, we investigate non-parametric Bayesian mixtures which are based on general flexible processes such as the Dirichlet process and the Chinese Restaurant Process. This non-parametric model formulation is relevant for both learning the model, as well for dealing with the issue of model selection. We propose new Bayesian non-parametric parsimonious mixtures and derive a MCMC sampling technique where the mixture model and the number of mixture components are simultaneously learned from the data. The selection of the model structure is performed by using Bayes Factors. These models, by their non-parametric and sparse formulation, are useful for the analysis of large data sets when the number of classes is undetermined and increases with the data, and when the dimension is high. The models are validated on simulated data and standard real data sets. Then, they are applied to a real difficult problem of automatic structuring of complex bioacoustic data issued from whale song signals. Finally, we open Markovian perspectives via hierarchical Dirichlet processes hidden Markov models
Prendes, Jorge. „New statistical modeling of multi-sensor images with application to change detection“. Thesis, Université Paris-Saclay (ComUE), 2015. http://www.theses.fr/2015SACLC006/document.
Der volle Inhalt der QuelleRemote sensing images are images of the Earth surface acquired from satellites or air-borne equipment. These images are becoming widely available nowadays and its sensor technology is evolving fast. Classical sensors are improving in terms of resolution and noise level, while new kinds of sensors are proving to be useful. Multispectral image sensors are standard nowadays and synthetic aperture radar (SAR) images are very popular.The availability of different kind of sensors is very advantageous since it allows us to capture a wide variety of properties of the objects contained in a scene. These properties can be exploited to extract richer information about these objects. One of the main applications of remote sensing images is the detection of changes in multitemporal datasets (images of the same area acquired at different times). Change detection for images acquired by homogeneous sensors has been of interest for a long time. However the wide range of different sensors found in remote sensing makes the detection of changes in images acquired by heterogeneous sensors an interesting challenge.Accurate change detectors adapted to heterogeneous sensors are needed for the management of natural disasters. Databases of optical images are readily available for an extensive catalog of locations, but, good climate conditions and daylight are required to capture them. On the other hand, SAR images can be quickly captured, regardless of the weather conditions or the daytime. For these reasons, optical and SAR images are of specific interest for tracking natural disasters, by detecting the changes before and after the event.The main interest of this thesis is to study statistical approaches to detect changes in images acquired by heterogeneous sensors. Chapter 1 presents an introduction to remote sensing images. It also briefly reviews the different change detection methods proposed in the literature. Additionally, this chapter presents the motivation to detect changes between heterogeneous sensors and its difficulties.Chapter 2 studies the statistical properties of co-registered images in the absence of change, in particular for optical and SAR images. In this chapter a finite mixture model is proposed to describe the statistics of these images. The performance of classical statistical change detection methods is also studied by taking into account the proposed statistical model. In several situations it is found that these classical methods fail for change detection.Chapter 3 studies the properties of the parameters associated with the proposed statistical mixture model. We assume that the model parameters belong to a manifold in the absence of change, which is then used to construct a new similarity measure overcoming the limitations of classic statistical approaches. Furthermore, an approach to estimate the proposed similarity measure is described. Finally, the proposed change detection strategy is validated on synthetic images and compared with previous strategies.Chapter 4 studies Bayesian non parametric algorithm to improve the estimation of the proposed similarity measure. This algorithm is based on a Chinese restaurant process and a Markov random field taking advantage of the spatial correlations between adjacent pixels of the image. This chapter also defines a new Jeffreys prior for the concentration parameter of this Chinese restaurant process. The estimation of the different model parameters is conducted using a collapsed Gibbs sampler. The proposed strategy is validated on synthetic images and compared with the previously proposed strategy. Finally, Chapter 5 is dedicated to the validation of the proposed change detection framework on real datasets, where encouraging results are obtained in all cases. Including the Bayesian non parametric model into the change detection strategy improves change detection performance at the expenses of an increased computational cost
Dallaire, Patrick. „Bayesian nonparametric latent variable models“. Doctoral thesis, Université Laval, 2016. http://hdl.handle.net/20.500.11794/26848.
Der volle Inhalt der QuelleOne of the important problems in machine learning is determining the complexity of the model to learn. Too much complexity leads to overfitting, which finds structures that do not actually exist in the data, while too low complexity leads to underfitting, which means that the expressiveness of the model is insufficient to capture all the structures present in the data. For some probabilistic models, the complexity depends on the introduction of one or more latent variables whose role is to explain the generative process of the data. There are various approaches to identify the appropriate number of latent variables of a model. This thesis covers various Bayesian nonparametric methods capable of determining the number of latent variables to be used and their dimensionality. The popularization of Bayesian nonparametric statistics in the machine learning community is fairly recent. Their main attraction is the fact that they offer highly flexible models and their complexity scales appropriately with the amount of available data. In recent years, research on Bayesian nonparametric learning methods have focused on three main aspects: the construction of new models, the development of inference algorithms and new applications. This thesis presents our contributions to these three topics of research in the context of learning latent variables models. Firstly, we introduce the Pitman-Yor process mixture of Gaussians, a model for learning infinite mixtures of Gaussians. We also present an inference algorithm to discover the latent components of the model and we evaluate it on two practical robotics applications. Our results demonstrate that the proposed approach outperforms, both in performance and flexibility, the traditional learning approaches. Secondly, we propose the extended cascading Indian buffet process, a Bayesian nonparametric probability distribution on the space of directed acyclic graphs. In the context of Bayesian networks, this prior is used to identify the presence of latent variables and the network structure among them. A Markov Chain Monte Carlo inference algorithm is presented and evaluated on structure identification problems and as well as density estimation problems. Lastly, we propose the Indian chefs process, a model more general than the extended cascading Indian buffet process for learning graphs and orders. The advantage of the new model is that it accepts connections among observable variables and it takes into account the order of the variables. We also present a reversible jump Markov Chain Monte Carlo inference algorithm which jointly learns graphs and orders. Experiments are conducted on density estimation problems and testing independence hypotheses. This model is the first Bayesian nonparametric model capable of learning Bayesian learning networks with completely arbitrary graph structures.
Roget-Vial, Céline. „deux contributions à l'étude semi-paramétrique d'un modèle de régression“. Phd thesis, Université Rennes 1, 2003. http://tel.archives-ouvertes.fr/tel-00008730.
Der volle Inhalt der QuelleViallon, Vivian. „Processus empiriques, estimation non paramétrique et données censurées“. Phd thesis, Université Pierre et Marie Curie - Paris VI, 2006. http://tel.archives-ouvertes.fr/tel-00119260.
Der volle Inhalt der QuelleDellagi, Hatem. „Estimations paramétrique et non paramétrique des données manquantes : application à l'agro-climatologie“. Paris 6, 1994. http://www.theses.fr/1994PA066546.
Der volle Inhalt der QuelleElamine, Abdallah Bacar. „Régression non-paramétrique pour variables fonctionnelles“. Thesis, Montpellier 2, 2010. http://www.theses.fr/2010MON20017.
Der volle Inhalt der QuelleThis thesis is divided in four sections with an additionnal presentation. In the first section, We expose the essential mathematics skills for the comprehension of the next sections. In the second section, we adress the problem of local non parametric with functional inputs. First, we propose an estimator of the unknown regression function. The construction of this estimator is related to the resolution of a linear inverse problem. Using a classical method of decomposition, we establish a bound for the mean square error (MSE). This bound depends on the small ball probability of the regressor which is assumed to belong to the class of Gamma varying functions. In the third section, we take again the work done in the preceding section by being situated in the frame of data belonging to a semi-normed space with infinite dimension. We establish bound for the MSE of the regression operator. This MSE can be seen as a function of the small ball probability function. In the last section, we interest to the estimation of the auxiliary function. Then, we establish the convergence in mean square and the asymptotic normality of the estimator. At last, by simulations, we study the bahavour of this estimator in a neighborhood of zero
Agbodan, Dago. „Nomination persistante dans un modèle paramétrique : identification non-ambigue͏̈ et appariement générique d'entités topologiques“. Poitiers, 2002. http://www.theses.fr/2002POIT2313.
Der volle Inhalt der QuelleParametric models have a dual structure where an abstract representation (the parametric specification) references an explicit representation (the geometry). The persistent naming problem is to maintain the references between these two representations in order to be able to reevaluate the second starting from the first, in spite of modifications. The problem is to identify an entity in an initial model, then to find it in a reevaluated model. We propose to represent evolutions of the shells and faces of the modeled objects in a graph. Each entity referenced by the specification is characterized in terms of the graph nodes, and by a link to the current geometry. Matching the initial graph and a reevaluated graph throughout a revaluation, and then, searching common elements in these graphs, allows us to interpret the references and thus to maintain the link between the parametric specification and the geometry in the reevaluated object, ensuring a persistent naming
Maillou, Balbine. „Caractérisation et identification non-paramétrique des non-linéarités de suspensions de haut-parleurs“. Thesis, Le Mans, 2015. http://www.theses.fr/2015LEMA1028.
Der volle Inhalt der QuelleThis thesis deals with the low frequencies mechanical behavior of the electrodynamic loudspeaker moving part, and especially with the suspensions, whose properties are among the most difficult to identify because of both assembly geometry and intrinsic materials, leading to nonlinear viscoelastic behaviors. In small signal domain, the Thiele and Small model describes the behavior of the whole loudspeaker with a good fit, the moving part behavior being modeled by a simple linear mass-spring system, with mass, damping and stiffness parameters. In large-signal domain, this model is no longer sufficient. Our approach is then to perform nonlinear system identification as a tool helping to improve analytical models. A model without physical knowledge is chosen : « Generalized Hammerstein ». Its identification requires the acquisition of experimental signals. A multi sensor experimental set up were so carried out and allows to characterize the whole moving part of a loudspeaker, without magnetic motor, attached to a rigid stand and excited with high axial displacement values, by means of a shaker. Shaker being itself a nonlinear device, a new method of « Generalized Hammerstein » model identification was developped, dedicated to nonlinear systems in series. Finally, parameters of an «expanded Thiele and Small» model are derived from the «Generalized Hammerstein» model parameters. This allows to highlight the evolution of the stiffness and damping with the frequency of excitation, with the displacement of the membrane, as well as the dependence of observed phenomena with the excitation level
Sansonnet, Laure. „Inférence non-paramétrique pour des interactions poissoniennes“. Phd thesis, Université Paris Sud - Paris XI, 2013. http://tel.archives-ouvertes.fr/tel-00835427.
Der volle Inhalt der QuelleVincent, Thomas. „Modèles hémodynamiques spatiaux adaptatifs pour l'imagerie cérébrale fonctionnelle“. Paris 11, 2010. http://www.theses.fr/2010PA112365.
Der volle Inhalt der QuelleThe approaches developed in this PhD take place in the analysis of functional brain imaging seeking the characterization of brain structures specialization. The central modality was functional magnetic resonance imaging (fMRI) which provides an indirect, hemodynamic, measure of the neural activity. Data analysis methods are conventionally divided into: (i) a localization task of activations and (ii) an estimation task i. E. Characterizing the hemodynamic response function (HRF) linking the stimulations provided by the paradigm to the observed fMRI signal. This PhD addresses the tasks (i) and (ii) simultaneously in a joint detection-estimation model (JDE), respecting the obvious interdependence of these two processes. The JDE approach here has been extended to express a model of spatial correlation on the local response level associated with the HRF, enabling the approach to be multivariate for the detection as well as the estimation tasks. In the Bayesian framework, this modeling is achieved by the expression of a prior discrete Markov field involving a regularization factor. The unsupervised treatment regarding this parameter for the whole brain has been developed by adaptively taking into account the heterogeneity of the geometries of brain regions. The approach is validated on the cortical surface, but also in the volume through several group analyses with different acquisition conditions. These were used to assess the impact of the method in terms of significance of activation and its positioning relative to the traditional approach
Taupin, Marie-Luce. „Estimation semi-paramétrique pour le modèle de régression non linéaire avec erreurs sur les variables“. Paris 11, 1998. http://www.theses.fr/1998PA112004.
Der volle Inhalt der QuelleGauzère, Franck. „Approche non-paramétrique pour un modèle 3 états avec censures par intervalles : application à la dépendance“. Bordeaux 2, 2000. http://www.theses.fr/2000BOR28709.
Der volle Inhalt der QuelleGuin, Ophélie. „Méthodes bayésiennes semi-paramétriques d'extraction et de sélection de variables dans le cadre de la dendroclimatologie“. Phd thesis, Université Paris Sud - Paris XI, 2011. http://tel.archives-ouvertes.fr/tel-00636704.
Der volle Inhalt der QuelleViallefont, Valérie. „Analyses bayesiennes du choix de modèles en épidémiologie : sélection de variables et modélisation de l'hétérogénéité pour des évènements“. Paris 11, 2000. http://www.theses.fr/2000PA11T023.
Der volle Inhalt der QuelleThis dissertation has two separated parts. In the first part, we compare different strategies for variable selection in a multivariate logistic regression model. Covariate and confounder selection in case-control studies is often carried out using either a two-step method or a stepwise variable selection method. Inference is then carried out conditionally on the selected model, but this ignores the madel uncertainty implicit in the variable selection process, and so underestimates uncertainty about relative risks. It is well known, and showed again in our study, that the ρ-values computed after variable selection can greatly overstate the strength of conclusions. We propose Bayesian Model Averaging as a formal way of taking account of madel uncertainty in a logistic regression context. The BMA methods, that allows to take into account several models, each being associated with its posterior probability, yields an easily interpreted summary, the posterior probability that a variable is a risk factor, and its estimate averaged over the set of models. We conduct two comparative simulations studies : the first one has a simple design including only one risk factor and one confounder, the second one mimics a epidemiological cohort study dataset, with a large number of potential risk factors. Our criteria are the mean bias, the rate of type I and type II errors, and the assessment of uncertainty in the results, which is bath more accurate and explicit under the BMA analysis. The methods are applied and compared in the context of a previously published case-control study of cervical cancer. The choice of the prior distributions are discussed. In the second part, we focus on the modelling of rare events via a Poisson distribution, that sometimes reveals substantial over-dispersion, indicating that sorme un explained discontinuity arises in the data. We suggest to madel this over-dispersion by a Poisson mixture. In a hierarchical Bayesian model, the posterior distributions of he unknown quantities in the mixture (number of components, weights, and Poisson parameters) can be estimated by MCMC algorithms, including reversible jump algothms which allows to vary the dimension of the mixture. We focus on the difficulty of finding a weakly informative prior for the Poisson parameters : different priors are detailed and compared. Then, the performances of different maves created for changing dimension are investigated. The model is extended by the introduction of covariates, with homogeneous or heterogeneous effect. Simulated data sets are designed for the different comparisons, and the model is finally illustrated in two different contexts : an ecological analysis of digestive cancer mortality along the coasts of France, and a dataset concerning counts of accidents in road-junctions
Moumouni, Kairou. „Etude et conception d'un modèle mixte sémiparamétrique stochastique pour l'analyse des données longitudinales environnementales“. Rennes 2, 2005. http://www.theses.fr/2005REN20052.
Der volle Inhalt der QuelleThis thesis is dealing with the analysis of longitudinal data that can be encountered in environmental studies. The general approach is based on the stochastic linear mixed model, that we extend using semiparametric techniques, such as penalized cubic splines. First, estimation methods are developed for the semiparametric stochastic mixed model, and then a simulation study is performed to measure the performances of the parameter estimates. In a second part, we propose an extension of the Cook's local influence method, in order to produce a sensibility analysis of our model and detect the effect of the perturbation of the structural components of the model. Some asymptotic properties of the local influence matrix are exhibited. Finally, the proposed model is applied to two real datasets : first, the analysis of nitrate concentration measurements in different locations of a watershed ; second, the analysis of bacteriological pollution of coastal bathing waters
Lesquoy-de, Turckheim Élisabeth. „Tests non paramétriques et rééchantillonnage : le modèle de Cox périodique“. Paris 11, 1987. http://www.theses.fr/1987PA112474.
Der volle Inhalt der QuelleThe first part proposes two nonparametric test defined by a simulation. One compares two distributions functions in a two-by-two black design, the other tests the independence of two censored survival times. The second part is an adaptation of Cox's regression model to a counting process having a periodic underlying intensity and predictable processes as regressors. These processes are ergodic and ϕ-mixing. The underlying intensity is estimated using either an empirical distribution-type estimate or a histogram-type estimate. These two estimates are asymptotically Gaussian and equivalent, as well as the associated regression parameters estimates. Finally, the model is applied to the analysis of a feeding pattern. The third part is a. Modelling of the kinetics of drought rhizogenesis of Sinapis alba
Libengue, Dobele-kpoka Francial Giscard Baudin. „Méthode non-paramétrique des noyaux associés mixtes et applications“. Thesis, Besançon, 2013. http://www.theses.fr/2013BESA2007/document.
Der volle Inhalt der QuelleWe present in this thesis, the non-parametric approach using mixed associated kernels for densities withsupports being partially continuous and discrete. We first start by recalling the essential concepts of classical continuousand discrete kernel density estimators. We give the definition and characteristics of these estimators. We also recall thevarious technical for the choice of smoothing parameters and we revisit the problems of supports as well as a resolutionof the edge effects in the discrete case. Then, we describe a new method of continuous associated kernels for estimatingdensity with bounded support, which includes the classical continuous kernel method. We define the continuousassociated kernels and we propose the mode-dispersion for their construction. Moreover, we illustrate this on the nonclassicalassociated kernels of literature namely, beta and its extended version, gamma and its inverse, inverse Gaussianand its reciprocal, the Pareto kernel and the kernel lognormal. We subsequently examine the properties of the estimatorswhich are derived, specifically, the bias, variance and the pointwise and integrated mean squared errors. Then, wepropose an algorithm for reducing bias that we illustrate on these non-classical associated kernels. Some simulationsstudies are performed on three types of estimators lognormal kernels. Also, we study the asymptotic behavior of thecontinuous associated kernel estimators for density. We first show the pointwise weak and strong consistencies as wellas the asymptotic normality. Then, we present the results of the global weak and strong consistencies using uniform andL1norms. We illustrate this on three types of lognormal kernels estimators. Subsequently, we study the minimaxproperties of the continuous associated kernel estimators. We first describe the model and we give the technicalassumptions with which we work. Then we present our results that we apply on some non-classical associated kernelsmore precisely beta, gamma and lognormal kernel estimators. Finally, we combine continuous and discrete associatedkernels for defining the mixed associated kernels. Using the tools of the unification of discrete and continuous analysis,we show the different properties of the mixed associated kernel estimators. All through this work, we choose thesmoothing parameter using the least squares cross-validation method
Le, Thi Xuan Mai. „Estimation semi-paramétrique et application à l’évaluation de la biomasse d'anchois“. Thesis, Toulouse, INSA, 2010. http://www.theses.fr/2010ISAT0006/document.
Der volle Inhalt der QuelleThe motivation of this study is to evaluate the anchovy biomass, that is estimate the egg densities at the spawning time and the mortality rate. The data are the anchovy egg densities that are the egg weights by area unit, collected in the Gascogne bay. The problem we are faced is to estimate from these data the egg densities at the spawning time. Until now, this is done by using the classical exponential mortality model. However, such model is inadequate for the data under consideration because of the great spatial variability of the egg densities at the spawning time. They are samples of generated by a r.v whose mathematical expectation is a0 and the probability density function is fA. Therefore, we propose an extended exponential mortality model Y (tj,kj) = A (tj,kj) e-z0tj +e(tj,kj) where A(tj,kj) and e(tj,kj) are i.i.d, with the random variables A and e being assumed to be independent. Then the problem consists in estimating the mortality rate and the probability density of the random variable . We solve this semiparametric estimation problem in two steps. First, we estimate the mortality rate by fitting an exponential mortality model to averaged data. Second, we estimate the density fA by combining nonparametric estimation method with deconvolution technique and estimate the parameter z0. Theoretical results of consistence of these estimates are corroborated by simulation studies
Gendre, Xavier. „Estimation par sélection de modèle en régression hétéroscédastique“. Phd thesis, Université de Nice Sophia-Antipolis, 2009. http://tel.archives-ouvertes.fr/tel-00397608.
Der volle Inhalt der QuelleLa première partie de cette thèse consiste dans l'étude du problème d'estimation de la moyenne et de la variance d'un vecteur gaussien à coordonnées indépendantes. Nous proposons une méthode de choix de modèle basée sur un critère de vraisemblance pénalisé. Nous validons théoriquement cette approche du point de vue non-asymptotique en prouvant des majorations de type oracle du risque de Kullback de nos estimateurs et des vitesses de convergence uniforme sur les boules de Hölder.
Un second problème que nous abordons est l'estimation de la fonction de régression dans un cadre hétéroscédastique à dépendances connues. Nous développons des procédures de sélection de modèle tant sous des hypothèses gaussiennes que sous des conditions de moment. Des inégalités oracles non-asymptotiques sont données pour nos estimateurs ainsi que des propriétés d'adaptativité. Nous appliquons en particulier ces résultats à l'estimation d'une composante dans un modèle de régression additif.
Vitse, Matthieu. „Réduction de modèle pour l'analyse paramétrique de l'endommagement dans les structures en béton armé“. Thesis, Université Paris-Saclay (ComUE), 2016. http://www.theses.fr/2016SACLN055/document.
Der volle Inhalt der QuelleThis thesis is dedicated to the development of an algorithm for the resolution of nonlinear problems for which there is a variability on some of the model parameters or on the loading conditions, which are only described by their intervals of variation. This study is part of the SINAPS@ project, which aims at evaluating the uncertainties in civil engineering structures and to quantify their influence on the global mechanical response of a structure to a seismic hazard. Unlike statistical or probabilistic approaches, we rely here on a deterministic approach. However, in order to reduce the computation cost of such problems, a PGD-based reduced-order modeling approach is implemented, for which the uncertain parameters are considered as additional variables of the problem. This method was implemented into the LATIN algorithm, which uses an iterative approach to solve the nonlinear aspect of the equations of the mechanical problem. This work present the extension of the classical time-space LATIN—PGD algorithm to parametric problems for which the parameters are considered as additional variables in the definition of the quantities of interest, as well as the application of such method to a damage model with unilateral effect, highlighting a variability on both material parameters and the loading amplitude. The feasibility of such coupling is illustrated on numerical examples for reinforced concrete structures subjected to different types of cyclic loading conditions (tension—compression, bending)
Tran, Gia-Lac. „Advances in Deep Gaussian Processes : calibration and sparsification“. Electronic Thesis or Diss., Sorbonne université, 2020. https://accesdistant.sorbonne-universite.fr/login?url=https://theses-intra.sorbonne-universite.fr/2020SORUS410.pdf.
Der volle Inhalt der QuelleGaussian Processes (GPs) are an attractive specific way of doing non-parametric Bayesian modeling in a supervised learning problem. It is well-known that GPs are able to make inferences as well as predictive uncertainties with a firm mathematical background. However, GPs are often unfavorable by the practitioners due to their kernel's expressiveness and the computational requirements. Integration of (convolutional) neural networks and GPs are a promising solution to enhance the representational power. As our first contribution, we empirically show that these combinations are miscalibrated, which leads to over-confident predictions. We also propose a novel well-calibrated solution to merge neural structures and GPs by using random features and variational inference techniques. In addition, these frameworks can be intuitively extended to reduce the computational cost by using structural random features. In terms of computational cost, the exact Gaussian Processes require the cubic complexity to training size. Inducing point-based Gaussian Processes are a common choice to mitigate the bottleneck by selecting a small set of active points through a global distillation from available observations. However, the general case remains elusive and it is still possible that the required number of active points may exceed a certain computational budget. In our second study, we propose Sparse-within-Sparse Gaussian Processes which enable the approximation with a large number of inducing points without suffering a prohibitive computational cost
Lacour, Claire. „Estimation non paramétrique adaptative pour les chaînes de Markov et les chaînes de Markov cachées“. Phd thesis, Université René Descartes - Paris V, 2007. http://tel.archives-ouvertes.fr/tel-00180107.
Der volle Inhalt der QuelleClertant, Matthieu. „Semi-parametric bayesian model, applications in dose finding studies“. Thesis, Paris 6, 2016. http://www.theses.fr/2016PA066230/document.
Der volle Inhalt der QuellePhase I clinical trials is an area in which statisticians have much to contribute. For over 30 years, this field has benefited from increasing interest on the part of statisticians and clinicians alike and several methods have been proposed to manage the sequential inclusion of patients to a study. The main purpose is to evaluate the occurrence of dose limiting toxicities for a selected group of patients with, typically, life threatening disease. The goal is to maximize the potential for therapeutic success in a situation where toxic side effects are inevitable and increase with increasing dose. From a range of given doses, we aim to determine the dose with a rate of toxicity as close as possible to some threshold chosen by the investigators. This dose is called the MTD (maximum tolerated dose). The standard situation is where we have a finite range of doses ordered with respect to the probability of toxicity at each dose. In this thesis we introduce a very general approach to modeling the problem - SPM (semi-parametric methods) - and these include a large class of methods. The viewpoint of SPM allows us to see things in, arguably, more relevant terms and to provide answers to questions such as asymptotic behavior. What kind of behavior should we be aiming for? For instance, can we consistently estimate the MTD? How, and under which conditions? Different parametrizations of SPM are considered and studied theoretically and via simulations. The obtained performances are comparable, and often better, to those of currently established methods. We extend the findings to the case of partial ordering in which more than one drug is under study and we do not necessarily know how all drug pairs are ordered. The SPM model structure leans on a hierarchical set-up whereby certain parameters are linearly constrained. The theoretical aspects of this structure are outlined for the case of distributions with discrete support. In this setting the great majority of laws can be easily considered and this enables us to avoid over restrictive specifications than can results in poor behavior
Caouder, Nathalie. „Régression non-linéaire paramétrique : etude de méthodes pour détecter des écarts au modèle. Maquette de système expert pour l'estimation des paramètres“. Paris 7, 1993. http://www.theses.fr/1993PA077132.
Der volle Inhalt der QuelleBatou, Anas. „Identification des forces stochastiques appliquées à un système dynamique non linéaire en utilisant un modèle numérique incertain et des réponses expérimentales“. Phd thesis, Université Paris-Est, 2008. http://tel.archives-ouvertes.fr/tel-00472080.
Der volle Inhalt der QuellePoilleux-Milhem, Hélène. „Test de validation adaptatif dans un modèle de régression : modélisation et estimation de l'effet d'une discontinuité du couvert végétal sur la dispersion du pollen de colza“. Paris 11, 2002. http://www.theses.fr/2002PA112297.
Der volle Inhalt der QuelleThis thesis framework is the spread of genetically modified organisms in the environment. Several parametric models of the individual pollen dispersal distribution have already been proposed for homogeneous experiments (plants emitting marked pollen surrounded by the same unmarked plants). In order to predict the "genetic pollution" in an agricultural landscape, a discontinuity effect on pollen flows in a cultivated area (e. G. A road crosses a field) has to be taken into account. This effect was modelled and estimated: according to the size of the discontinuity, it may correspond to a significant acceleration of the pollen flow. Graphical diagnosis methods show that the modelling of the individual pollen dispersal distribution and of the discontinuity effect, is best fitting the data when using constant piecewise functions. Prior to using parametric models to predict genetic pollution, goodness-of-fit tools are essential. We therefore propose a goodness-of-fit test in a nonlinear Gaussian regression model, where the errors are independent and identically distributed. This test does not require any knowledge on the regression function and on the variance of the observations. It generalises the linear hypothesis tests proposed by Baraud et al (Ann. Statist. 2003, Vol. 31) to the nonlinear hypothesis. It is asymptotically of level α and a set of functions over which it is asymptotically powerful is characterized. It is rate optimal among adaptive procedures over isotropic and anisotropic Hölder classes of alternatives. It is consistent against directional alternatives that approach the null hypothesis at a rate close to the parametric rate. According to a simulation study, this test is powerful even for fixed sample sizes
Verdière, Nathalie. „Identifiabilité de systèmes d'équations aux dérivées partielles semi-discrétisées et applications à l'identifiabilité paramétrique de modèles en pharmacocinétique et en pollution“. Phd thesis, Université de Technologie de Compiègne, 2005. http://tel.archives-ouvertes.fr/tel-00011838.
Der volle Inhalt der QuelleDans cette thèse, deux modèles non linéaires en pharmacocinétique de type Michaelis-Menten ont tout d'abord été étudiés. Ensuite, nous nous sommes intéressés à un modèle de pollution décrit par une équation aux dérivées partielles parabolique. Le terme source à identifier était modélisé par le produit de la fonction débit avec la masse de Dirac, de support la position de la source polluante. Le but du travail était de fournir une première estimation de la source polluante. Après avoir obtenu l'identifiabilité du problème continu, nous avons étudié l'identifiabilité d'un problème approché en nous appuyant sur les méthodes d'algèbre différentielle. Celui-ci a été obtenu en approchant la masse de Dirac par une fonction gaussienne et en discrétisant ensuite le système en espace. Les résultats d'identifiabilité ont été obtenus quel que soit le nombre de points de discrétisation en espace. De cette étude théorique, nous en avons déduit des algorithmes numériques donnant une première estimation des paramètres à identifier.
Godard, Pierre. „Unsupervised word discovery for computational language documentation“. Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLS062/document.
Der volle Inhalt der QuelleLanguage diversity is under considerable pressure: half of the world’s languages could disappear by the end of this century. This realization has sparked many initiatives in documentary linguistics in the past two decades, and 2019 has been proclaimed the International Year of Indigenous Languages by the United Nations, to raise public awareness of the issue and foster initiatives for language documentation and preservation. Yet documentation and preservation are time-consuming processes, and the supply of field linguists is limited. Consequently, the emerging field of computational language documentation (CLD) seeks to assist linguists in providing them with automatic processing tools. The Breaking the Unwritten Language Barrier (BULB) project, for instance, constitutes one of the efforts defining this new field, bringing together linguists and computer scientists. This thesis examines the particular problem of discovering words in an unsegmented stream of characters, or phonemes, transcribed from speech in a very-low-resource setting. This primarily involves a segmentation procedure, which can also be paired with an alignment procedure when a translation is available. Using two realistic Bantu corpora for language documentation, one in Mboshi (Republic of the Congo) and the other in Myene (Gabon), we benchmark various monolingual and bilingual unsupervised word discovery methods. We then show that using expert knowledge in the Adaptor Grammar framework can vastly improve segmentation results, and we indicate ways to use this framework as a decision tool for the linguist. We also propose a tonal variant for a strong nonparametric Bayesian segmentation algorithm, making use of a modified backoff scheme designed to capture tonal structure. To leverage the weak supervision given by a translation, we finally propose and extend an attention-based neural segmentation method, improving significantly the segmentation performance of an existing bilingual method
Coudin, Élise. „Inférence exacte et non paramétrique dans les modèles de régression et les modèles structurels en présence d'hétéroscédasticité de forme arbitraire“. Thèse, Paris, EHESS, 2007. http://hdl.handle.net/1866/1506.
Der volle Inhalt der QuelleLe, Corff Sylvain. „Estimations pour les modèles de Markov cachés et approximations particulaires : Application à la cartographie et à la localisation simultanées“. Phd thesis, Telecom ParisTech, 2012. http://tel.archives-ouvertes.fr/tel-00773405.
Der volle Inhalt der QuelleKoladjo, Babagnidé François. „Estimation non paramétrique du nombre d'espèces : Application à l'étude de la faune ichtyologique du bassin du fleuve Ouëmé“. Thesis, Paris 11, 2013. http://www.theses.fr/2013PA112153.
Der volle Inhalt der QuelleThis manuscript is structured in two parts. The #rst part composed of Chapters 2to 4 deals with the problem of estimating the number of classes in a population withan application in ecology. The second part, corresponding to Chapter 5, concernsthe application of statistical methods to analyze fisheries data.In the first part, we consider a heterogeneous population split into several classes.From a sample, the numbers of observed individuals per class, also called abun-dances, are used to estimate the total number of classes in the population. In theliterature devoted to the number of classes estimation, methods based on a mix-ture of Poisson distributions seem to be the most effcient (see for example the workof Chao and Bunge (2002) in the parametric framework and that of Wang and Lind-say (2005) in a non-parametric framework). Applications of these approaches to realdata show that the distribution of abundances can be approximated by a convexdistribution. We propose a non-parametric approach to estimate the distribution ofabundances under the constraint of convexity. This constraint defines a theoreticalframework for estimating a discrete density. The problem of estimating the numberof classes is then tackled in two steps.We show on the one hand the existence and uniqueness of an estimator of adiscrete density under the constraint of convexity. Under this constraint, we provethat a discrete density can be written as a mixture of triangular distributions. Usingthe support reduction algorithm proposed by Groeneboom et al. (2008), we proposean exact algorithm to estimate the proportions in the mixture.On the other hand, the estimation procedure of a discrete convex density is usedto estimate the zero-truncated distribution of the observed abundance data. Thezero-truncated distribution estimate is then extended at zero to derive an estimateof the probability that a class is not observed. This extension is made so as tocancel the first component in the mixture of triangular distributions. An estimateof the total number of classes is obtained through a binomial model assuming thateach class appears in a sample by a Bernoulli trial. We show the convergence inlaw of the proposed estimator. On practical view, an application to real ecologicaldata is presented. The method is then compared to other concurrent methods usingsimulations.The second part presents the analysis of fisheries data collected on the Ouémériver in Benin. We propose a statistical approach for grouping species accordingto their temporal abundance profile, to estimate the stock of a species and theircatchability by artisanal fishing gears
Dalalyan, Arnak. „Contribution à la statistique des diffusions. Estimation semiparamétrique et efficacité au second ordre.Agrégation et réduction de dimension pour le modèle de régression“. Habilitation à diriger des recherches, Université Pierre et Marie Curie - Paris VI, 2007. http://tel.archives-ouvertes.fr/tel-00192080.
Der volle Inhalt der QuelleLe premier chapitre contient une description générale des résultats obtenus en les replaçant dans un contexte historique et en présentant les motivations qui nous ont animées pour étudier ces problèmes. J'y décris également de façon informelle les idées clés des démonstrations.
Au second chapitre, je présente les définitions principales nécessaires pour énoncer de façon rigoureuse les résultats les plus importants. Ce chapitre contient également une discussion plus formelle permettant de mettre en lumière certains aspects théoriques et pratiques de nos résultats.
Antic, Julie. „Méthodes non-paramétriques en pharmacocinétique et/ou pharmacodynamie de population“. Toulouse 3, 2009. http://thesesups.ups-tlse.fr/935/.
Der volle Inhalt der QuelleThis thesis studies non-parametric (NP) methods for the estimation of random-effects' distribution in non-linear mixed effect models. The objective is to evaluate the interest of these methods for population Pharmacokinetics (PK) and/or Pharmacodynamics (PD) analyses within Pharmaceutical industry. In a first step, the thesis reviews the statistical properties of four important NP methods. Besides, their practical performances are evaluated using some simulation studies, inspired from population PK analyses. The interest of NP methods is established in theory and in practice. NP methods are then for the population PK/PD analysis of an anti-diabetic drug. The aim is to evaluate the methods abilities to detect a sub-population of nonresponder patients. Some simulation studies show that two NP methods seem more capable of detecting this sub-population. The last part of the thesis is dedicated to the research of stochastic algorithms that improve the computation of NP methods. A perturbed stochastic gradient algorithm is proposed