-
News
- Journée au vert POLARIS 2022/05/23
- DATAMOVE/POLARIS picnic 2021/06/22
- DATAMOVE/POLARIS BBQ 2019 2019/06/14
- POLARIS Bootcamp (May 2019) 2019/05/24
- slides of Andras Gyorgy 2016/01/15
Next seminars
Events
Events in February–March 2023
MMonday TTuesday WWednesday TThursday FFriday SSaturday SSunday 30January 30, 202331January 31, 2023February
1February 1, 20232February 2, 20233February 3, 20234February 4, 20235February 5, 20236February 6, 2023Seminar Polaris-tt Learning in finite-horizon MDP with UCB (Romain Cravic)
–
February 7, 2023Most of you probably know Markov Decisions Processes (MDP). They are very useful to handle situations where an agent interacts with an environnement that may involve randomness. Concretely, at each time the MDP has a current state and the agent chooses an action : This couple state-action induces a (random) reward and a (random) state transition. If the probability distributions for rewards and transitions are known, at least theoretically, designing optimal behaviors for the agent is easy. What about the case where these distributions are unknown at the early stage of the process ? How to LEARN optimal behaviors efficiently ? A popular way to handle this issue is to use the optimism paradigm, inspired from UCB algorithms designed for stochastic bandits problems. In this talk, I will expose the main ideas of two possible approaches, UCRL algorithm and optimistic Q-learning algorithm, that use optimism to well perform in finite-horizon
Bâtiment IMAG (406)Saint-Martin-d'Hères, 38400France8February 8, 20239February 9, 202310February 10, 202311February 11, 202312February 12, 202313February 13, 202314February 14, 202315February 15, 202316February 16, 202317February 17, 202318February 18, 202319February 19, 202320February 20, 2023Seminar Polaris-tt: Decomposition of Normal Form Games - Harmonic, Potential, and Non-Strategic Games (Davide Legacci)
–
February 21, 2023In this talk, we will explore the concept of normal form games and their decomposition into non-strategic, harmonic, and potential games. We will begin by introducing the response graph of a game, which is a visual representation of the strategies available to each player and their corresponding utilities. What dictates the strategic interaction among players is the difference between utilities, rather than the utilities themselves. We will introduce an object that captures this behavior, called deviation flow of the game, and use it to define non-strategic, harmonic, and potential games. Finally, we will discuss the properties of these components.
Bâtiment IMAG (442)22February 22, 202323February 23, 202324February 24, 202325February 25, 202326February 26, 202327February 27, 202328February 28, 2023March
1March 1, 20232March 2, 20233March 3, 20234March 4, 20235March 5, 20236March 6, 20237March 7, 20238March 8, 20239March 9, 202310March 10, 202311March 11, 202312March 12, 202313March 13, 202314March 14, 202315March 15, 202316March 16, 202317March 17, 202318March 18, 202319March 19, 202320March 20, 202321March 21, 202322March 22, 202323March 23, 202324March 24, 202325March 25, 202326March 26, 202327March 27, 202328March 28, 202329March 29, 2023PhD defense Kimang Khun: Apprentissage par renforcement dans les systèmes dynamiques structurés
–
March 30, 2023Thèse supervisée par Nicolas GAST et Bruno GAUJAL.Bâtiment IMAG (amphitheater)Saint-Martin-d'Hères, 38400France31March 31, 2023April
1April 1, 20232April 2, 2023Meta