Algorithme EM r\'egularis\'e

Autor: Houdouin, Pierre, Jonkcheere, Matthieu, Pascal, Frederic

Expectation-Maximization (EM) algorithm is a widely used iterative algorithm for computing maximum likelihood estimate when dealing with Gaussian Mixture Model (GMM). When the sample size is smaller than the data dimension, this could lead to a singu

Externí odkaz: http://arxiv.org/abs/2307.01955

Apprentissage profond non supervisé fondé sur l'algorithme EM pour la segmentation du mouvement

Autor: Meunier, Etienne, Badoual, Anaïs, Bouthemy, Patrick

Publikováno v: RFIAP 2022-Reconnaissance des Formes, Image, Apprentissage et Perception
RFIAP 2022-Reconnaissance des Formes, Image, Apprentissage et Perception, Jul 2022, Vannes, France. pp.1-10

This paper presents a CNN-based fully unsupervised method for motion segmentation from optical flow. We assume that the input optical flow can be represented as a piecewise set of parametric motion models, typically, affine or quadratic motion models

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::dde7aa5b28bfb32fdcbbdd2b7ffa3dbf
https://hal.inria.fr/hal-03926935

Algorithme EM en ligne pour des données fonctionnelles

Autor: Eric Moulines, Sidonie Lefebvre, Florian Maire

Publikováno v: Computational Statistics and Data Analysis
Computational Statistics and Data Analysis, Elsevier, 2017, 111, pp.27-47. ⟨10.1016/j.csda.2017.01.006⟩

International audience; A novel approach to perform unsupervised sequential learning for functional data is proposed. The goal is to extract reference shapes (referred to as templates) from noisy, deformed and censored realizations of curves and imag

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::05a2ec1f1b1bec951dacd9d3419ab5d1
https://doi.org/10.1016/j.csda.2017.01.006

Estimation de modèles à équations structurelles par algorithme EM pour l’analyse longitudinale de la qualité de vie relative à la santé en cancérologie

Autor: Antoine Barbieri, Myriam Tami, Xavier Bry, Sophie Gourgou, Christian Lavergne, D. Azria, Caroline Mollevi

Publikováno v: EPICLIN 9 / 22èmes journées des statisticiens des Centres de Lutte Contre le Cancer (CLCC)
EPICLIN 9 / 22èmes journées des statisticiens des Centres de Lutte Contre le Cancer (CLCC), May 2015, Montpellier, France

Introduction En oncologie, la qualite de vie relative a la sante (QdV) est devenue un critere essentiel dans les essais cliniques mais son analyse longitudinale reste complexe et non standardisee. Un des freins conceptuels de ce critere est son aspec

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::f3388ddf6b6ba0b4cbe405a40a67ebe2
https://hal.archives-ouvertes.fr/hal-01207258v2/file/Abstract_Epiclin.pdf

Estimation par algorithme EM pour modèles à facteurs et à équations structurelles

Autor: Tami, Myriam, Bry, Xavier, Lavergne, Christian

Publikováno v: JDS 2014
JDS 2014 Rennes
JDS 2014 Rennes, Jun 2014, Rennes, France

National audience; Introduced in the 1970s by Jöreskog, structural equation models with factors allow to connect unobservable variables, called latent. Originally, the only parameter estimation method used on that models was analysis of covariance s

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=od______2592::34d1db2c7c20418bfeb95af761adef5d
https://hal.archives-ouvertes.fr/hal-01075781