[en] This dissertation presents a detailed investigation of the rho-estimation approach applied to dependent data. The work it contains establishes non-asymptotic deviation bounds with respect to a Hellinger-type loss in the most general context. From there, we obtain non-asymptotic oracle inequalities and robustness properties within the context of mixture models, hidden Markov models, and diffusion processes.
Disciplines :
Mathematics
Author, co-author :
LECESTRE, Alexandre ; University of Luxembourg > Faculty of Science, Technology and Medicine > Department of Mathematics > Team Yannick BARAUD
Language :
English
Title :
Robust estimation for possibly dependent observations: application to mixture and hidden Markov models
Defense date :
14 December 2023
Institution :
Unilu - University of Luxembourg [Faculty of Science, Technology and Medicine], Esch-sur-Alzette, Luxembourg
Degree :
Docteur en Mathématiques (DIP_DOC_0004_B)
Promotor :
BARAUD, Yannick ; University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Mathematics (DMATH)
President :
PODOLSKIJ, Mark ; University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Mathematics (DMATH)
Secretary :
GASSIA Elisabeth; Université Paris-Saclay [FR] > Institut de Mathématique d'Orsay
Jury member :
DÜMBGEN Lutz; UniBe - Universität Bern [CH] > Institut für Mathematische Statistik und Versicherungslehre (IMSV)
REISS Markus; Humboldt-Universität zu Berlin [DE] > Institut für Mathematik
European Projects :
H2020 - 811017 - SanDAL - ERA Chair in Mathematical Statistics and Data Science for the University of Luxembourg