Markov, Semi-Markov Models and Associated Fields (from Theory to Application and back)
1-4 juil. 2025
Sorbonne Université UPMC - Paris (France)
Markov chains, semi-Markov chains, and more broadly, Markov processes and their hidden variants represent a rich and versatile family of stochastic models with applications spanning a wide range of fields, including population dynamics, epidemic modeling, seismology, speech and activity recognition, and reliability analysis, to name just a few. Despite their widespread use, ongoing theoretical and algorithmic advancements are essential to tackle emerging challenges in real-world problems, such as those involving complex observations, multiple interacting hidden dynamics, or control scenarios with constraints. These models are explored from diverse perspectives across multiple methodological communities—such as process statistics, computational statistics, and optimal control—and applied domains, including signal processing, ecology, and medicine. As part of the ANR HSMM-INCA project consortium launched in January 2022 (https://groupes.renater.fr/wiki/hsmm-inca/public/axes_de_recherche), a national workshop was organized in 2023 to bring together the French statistical community engaged in these topics. The event served as a platform to share the latest developments, encompassing theoretical insights, modeling techniques, inference methods, decision-making strategies, and practical applications. It also fostered productive exchanges, particularly between doctoral students and senior researchers. The 2025 workshop aims to achieve two key objectives: expanding its scope to an international audience and broadening its thematic reach. Hidden Markov and semi-Markov models are often developed independently within non-statistical communities, such as signal processing and artificial intelligence. The upcoming workshop seeks to bridge these gaps by fostering dialogue and collaboration across disciplines, particularly on topics at the intersection of these fields.
Discipline scientifique :
Statistiques
Lieu de la conférence