pr\u00e9sentent<\/p>\n\n\n\n
Autoris\u00e9e \u00e0 pr\u00e9senter ses travaux en vue de l\u2019obtention du Doctorat de l’Institut Polytechnique de Paris, pr\u00e9par\u00e9 \u00e0 T\u00e9l\u00e9com SudParis en :<\/p>\n\n\n\n
le\u00a0MERCREDI 2 OCTOBRE 2024\u00a0\u00e0 10h00 \u00e0 C06
Telecom SudParis 9 Rue Charles Fourier 91000 Evry-Courcouronnes<\/p>\n\n\n\n
Membres du jury :<\/strong><\/p>\n\n\n\n M. YOHAN PETETIN<\/strong>, Ma\u00eetre de conf\u00e9rences, T\u00e9l\u00e9com SudParis, FRANCE – Directeur de these R\u00e9sum\u00e9 :<\/strong><\/p>\n\n\n\n Cette th\u00e8se vise \u00e0 mod\u00e9liser des donn\u00e9es s\u00e9quentielles \u00e0 travers l’utilisation de mod\u00e8les probabilistes \u00e0 variables latentes et param\u00e9tr\u00e9s par des architectures de type r\u00e9seaux de neurones profonds. Notre objectif est de d\u00e9velopper des mod\u00e8les dynamiques capables de capturer des dynamiques temporelles complexes inh\u00e9rentes aux donn\u00e9es s\u00e9quentielles tout en \u00e9tant applicables dans des domaines vari\u00e9s tels que la classification, la pr\u00e9diction et la g\u00e9n\u00e9ration de donn\u00e9es pour n’importe quel type de donn\u00e9es s\u00e9quentielles. Notre approche se concentre sur plusieurs probl\u00e9matiques li\u00e9s \u00e0 la mod\u00e9lisation de ce type de donn\u00e9es, chacune \u00e9tant d\u00e9taill\u00e9 dans un chapitre de ce manuscrit. Dans un premier temps, nous balayons les principes fondamentaux de l’apprentissage profond et de l’estimation bay\u00e9sienne. Par la suite, nous nous focalisations sur la mod\u00e9lisation de donn\u00e9es s\u00e9quentielles par des mod\u00e8les de Markov cach\u00e9s qui constitueront le socle commun des mod\u00e8les g\u00e9n\u00e9ratifs d\u00e9velopp\u00e9s par la suite. Plus pr\u00e9cis\u00e9ment, notre travail s’int\u00e9resse au probl\u00e8me de la classification (bay\u00e9sienne) s\u00e9quentielle de s\u00e9ries temporelles dans diff\u00e9rents contextes : supervis\u00e9 (les donn\u00e9es observ\u00e9es sont \u00e9tiquet\u00e9es) ; semi-supervis\u00e9 (les donn\u00e9es sont partiellement \u00e9tiquet\u00e9es) ; et enfin non supervis\u00e9s (aucune \u00e9tiquette n’est disponible). Pour cela, la combinaison de r\u00e9seaux de neurones profonds avec des mod\u00e8les probabilistes markoviens vise \u00e0 am\u00e9liorer le pouvoir g\u00e9n\u00e9ratif des mod\u00e9lisations plus classiques mais pose de nombreux d\u00e9fis du point de vue de l’inf\u00e9rence bay\u00e9sienne : estimation d’un grand nombre de param\u00e8tres, estimation de lois \u00e0 post\u00e9riori et interpr\u00e9tabilit\u00e9 de certaines variables cach\u00e9es (les labels). En plus de proposer une solution pour chacun de ces probl\u00e8mes, nous nous int\u00e9ressons \u00e9galement \u00e0 des approches novatrices pour relever des d\u00e9fis sp\u00e9cifiques en imagerie m\u00e9dicale pos\u00e9s par le Groupe Europ\u00e9en de Recherche sur les Proth\u00e8ses Appliqu\u00e9es \u00e0 la Chirurgie Vasculaire (GEPROMED).<\/p>\n\n\n\n Abstract :<\/strong><\/p>\n\n\n\n This thesis explores and models sequential data by applying various probabilistic models with latent variables, complemented by deep neural networks. The motivation for this research is the development of dynamic models that adeptly capture the complex temporal dynamics inherent in sequential data. Designed to be versatile and adaptable, these models aim to be applicable across domains including classification, prediction, and data generation, and adaptable to diverse data types. The research focuses on several key areas, each detailed in its respective chapter. Initially, the fundamental principles of deep learning, and Bayesian estimation are introduced. Sequential data modeling is then explored, emphasizing the Markov chain models, which set the stage for the generative models discussed in subsequent chapters. In particular, the research delves into the sequential Bayesian classification of data in supervised, semi-supervised, and unsupervised contexts. The integration of deep neural networks with well-established probabilistic models is a key strategic aspect of this research, leveraging the strengths of both approaches to address complex sequential data problems more effectively. This integration leverages the capabilities of deep neural networks to capture complex nonlinear relationships, significantly improving the applicability and performance of the models. In addition to our contributions, this thesis also proposes novel approaches to address specific challenges posed by the Groupe Europ\u00e9en de Recherche sur les Proth\u00e8ses Appliqu\u00e9es \u00e0 la Chirurgie Vasculaire (GEPROMED). These proposed solutions reflect the practical and possible impactful application of this research, demonstrating its potential contribution to the field of vascular surgery.<\/p>\n","protected":false},"excerpt":{"rendered":" L’Ecole doctorale : Math\u00e9matiques Hadamard et le Laboratoire de recherche SAMOVAR – Services r\u00e9partis, Architectures, Mod\u00e9lisation, Validation, Administration des R\u00e9seaux pr\u00e9sentent l\u2019AVIS DE SOUTENANCE de Madame Katherine Tania MORALES QUINGA Autoris\u00e9e \u00e0 pr\u00e9senter ses travaux en vue de l\u2019obtention du Doctorat de l’Institut Polytechnique de Paris, pr\u00e9par\u00e9 \u00e0 T\u00e9l\u00e9com SudParis en : \u00ab Mod\u00e8les de […]<\/p>\n","protected":false},"author":4,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ocean_post_layout":"","ocean_both_sidebars_style":"","ocean_both_sidebars_content_width":0,"ocean_both_sidebars_sidebars_width":0,"ocean_sidebar":"","ocean_second_sidebar":"","ocean_disable_margins":"enable","ocean_add_body_class":"","ocean_shortcode_before_top_bar":"","ocean_shortcode_after_top_bar":"","ocean_shortcode_before_header":"","ocean_shortcode_after_header":"","ocean_has_shortcode":"","ocean_shortcode_after_title":"","ocean_shortcode_before_footer_widgets":"","ocean_shortcode_after_footer_widgets":"","ocean_shortcode_before_footer_bottom":"","ocean_shortcode_after_footer_bottom":"","ocean_display_top_bar":"default","ocean_display_header":"default","ocean_header_style":"","ocean_center_header_left_menu":"","ocean_custom_header_template":"","ocean_custom_logo":0,"ocean_custom_retina_logo":0,"ocean_custom_logo_max_width":0,"ocean_custom_logo_tablet_max_width":0,"ocean_custom_logo_mobile_max_width":0,"ocean_custom_logo_max_height":0,"ocean_custom_logo_tablet_max_height":0,"ocean_custom_logo_mobile_max_height":0,"ocean_header_custom_menu":"","ocean_menu_typo_font_family":"","ocean_menu_typo_font_subset":"","ocean_menu_typo_font_size":0,"ocean_menu_typo_font_size_tablet":0,"ocean_menu_typo_font_size_mobile":0,"ocean_menu_typo_font_size_unit":"px","ocean_menu_typo_font_weight":"","ocean_menu_typo_font_weight_tablet":"","ocean_menu_typo_font_weight_mobile":"","ocean_menu_typo_transform":"","ocean_menu_typo_transform_tablet":"","ocean_menu_typo_transform_mobile":"","ocean_menu_typo_line_height":0,"ocean_menu_typo_line_height_tablet":0,"ocean_menu_typo_line_height_mobile":0,"ocean_menu_typo_line_height_unit":"","ocean_menu_typo_spacing":0,"ocean_menu_typo_spacing_tablet":0,"ocean_menu_typo_spacing_mobile":0,"ocean_menu_typo_spacing_unit":"","ocean_menu_link_color":"","ocean_menu_link_color_hover":"","ocean_menu_link_color_active":"","ocean_menu_link_background":"","ocean_menu_link_hover_background":"","ocean_menu_link_active_background":"","ocean_menu_social_links_bg":"","ocean_menu_social_hover_links_bg":"","ocean_menu_social_links_color":"","ocean_menu_social_hover_links_color":"","ocean_disable_title":"default","ocean_disable_heading":"default","ocean_post_title":"","ocean_post_subheading":"","ocean_post_title_style":"","ocean_post_title_background_color":"","ocean_post_title_background":0,"ocean_post_title_bg_image_position":"","ocean_post_title_bg_image_attachment":"","ocean_post_title_bg_image_repeat":"","ocean_post_title_bg_image_size":"","ocean_post_title_height":0,"ocean_post_title_bg_overlay":0.5,"ocean_post_title_bg_overlay_color":"","ocean_disable_breadcrumbs":"default","ocean_breadcrumbs_color":"","ocean_breadcrumbs_separator_color":"","ocean_breadcrumbs_links_color":"","ocean_breadcrumbs_links_hover_color":"","ocean_display_footer_widgets":"default","ocean_display_footer_bottom":"default","ocean_custom_footer_template":"","ocean_post_oembed":"","ocean_post_self_hosted_media":"","ocean_post_video_embed":"","ocean_link_format":"","ocean_link_format_target":"self","ocean_quote_format":"","ocean_quote_format_link":"post","ocean_gallery_link_images":"on","ocean_gallery_id":[],"footnotes":""},"categories":[286,615],"tags":[],"class_list":["post-6581","post","type-post","status-publish","format-standard","hentry","category-fractualites-ennews-fr","category-seminaire-sop","entry"],"_links":{"self":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/6581","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/comments?post=6581"}],"version-history":[{"count":2,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/6581\/revisions"}],"predecessor-version":[{"id":6583,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/6581\/revisions\/6583"}],"wp:attachment":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/media?parent=6581"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/categories?post=6581"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/tags?post=6581"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
M. ERWAN LE PENNEC<\/strong>, Professeur, Ecole Polytechnique, FRANCE – Examinateur
M. STEPHANE DERRODE<\/strong>, Professeur, \u00c9cole Centrale de Lyon, FRANCE – Examinateur
Mme SYLVIE LE H\u00c9GARAT<\/strong>, Professeure, \u00c9cole normale sup\u00e9rieure Paris-Saclay, FRANCE – Examinateur
M. FRANCOIS SEPTIER<\/strong>, Professeur, Universite Bretagne Sud, FRANCE – Rapporteur
Mme Myriam MAUMY<\/strong>, Professeure des universit\u00e9s, \u00c9cole des hautes \u00e9tudes en sant\u00e9 publique (EHESP), FRANCE – Rapporteur<\/p>\n\n\n\n\u00ab Mod\u00e8les de Markov g\u00e9n\u00e9ratifs pour la classification s\u00e9quentielle bay\u00e9sienne \u00bb<\/h2>\n\n\n\n
pr\u00e9sent\u00e9 par Madame Katherine Tania MORALES QUINGA<\/h2>\n\n\n\n