8ème Journée COSMOS

8th GDT COSMOS workshop: "Stochastic Optimization and Reinforcement Learning"

Lieu
Ces Journées auront lieu à INRIA Paris (2 rue Simone Iff 75012 Paris).
Les informations Pour Venir à l'INRIA Paris.
La journée se déroulera en salle Lions 1 située au RDC (matinée et fin de l'après-midi) et en salle C434 située au 4eme étage (début de l'après-midi).
Attention l'accès à la salle C434 nécessite un badge (à se présenter à l'acceuil avec une pièce d'identité avec la photo).

Location
INRIA Paris (2 rue Simone Iff 75012 Paris).
How to get to INRIA Paris.
Rooms: Lions 1 on RDC level (morning and 2nd afternoon session) and room C434 on the 4th floor (1st afternoon session).
Warning: access to room C434 requires a badge (go to the information desk at the entrance of the building, you will need an ID document with photo to get access to the 4th floor).

Inscriptions
Cliquez sur l'onglet "register" ci dessus pour vous inscrire.
La participation à la journée est libre mais l'inscription est nécessaire pour des raisons pratiques.
Attention comme nous devons donner la liste des participants à l'accueil de l'INRIA. La clôture des inscriptions est le 6/11 à midi.

Registration
Please Click on the register tab above. Inscription are free but mandatory for practical purposes.
Be carefull that the registration will be closed at noon the 6th november.

Speakers

• Elene Anton. IRIT. Title: On the stability of redundancy models.
  Slides: Redundancy Models
• Pierre Coucheney. Laboratoire David, Université Versailles Saint Quentin.
  Title: Solving Simple Stochastic Games with Few Random Nodes Faster Using Bland's Rule.
  Slides: Simple Stochastic Games
• Pierre Gaillard. INRIA Paris (SIERRA).
  Title: Target Tracking for Contextual Bandits: Application to Demand Side Management.
  Slides: Contextual Bandits
• Erwan Le Pennec. CMAP, Ecole Polytechnique.
  Title: Optimization of a sequential decision problem in prenatal ultrasound.
  Slides: Optimization of sequential decisions
• Antonio Massaro. Nokia Bell Labs.
  Title: Optimal Trunk-Reservation by Policy Learning.
  Slides: Introduction gradient policy and Optimal Trunk-Reservation
• Sadegh Talebi. INRIA Lille (SEQUEL).
  Title: Efficient Regret Minimizing Strategies for Tabular Average-Reward MDPs.
  Slides: Tabular average reward MDP
• Pierre Henri Wuillemin. Lip6, Sorbonne Universités.
  Title: Factored Markov Decision Processes and Reinforcement Learning.

Program (subject to change)

•  9h00   Welcome and coffee (RDC)
•  9h30   Antonio Massaro "Optimal Trunk-Reservation by Policy Learning" - meeting room Lions 1 (RDC)
• 10h30   Pause (RDC)
• 11h00   Pierre Henri Wuillemin "Factored Markov Decision Processes and Reinforcement Learning" - meeting room Lions 1 (RDC)
• 11h45   Sadegh Talebi "Efficient Regret Minimizing Strategies for Tabular Average-Reward MDPs" - meeting room Lions 1 (RDC)
• 12h30   Lunch
• 14h00   Erwan Le Pennec "Optimization of a sequential decision problem in prenatal ultrasound" - meeting room C434 (4th floor)
• 14h45   Pierre Gaillard "Target Tracking for Contextual Bandits: Application to Demand Side Management" - meeting room C434 (4th floor)
• 15h30   Pause (RDC)
• 16h00   Elene Anton "On the stability of redundancy models"  meeting room Lions 1 (RDC)
• 16h45   Pierre Coucheney "Solving Simple Stochastic Games with Few Random Nodes Faster Using Bland's Rule" - meeting room Lions 1 (RDC)
• 17h30   Closing - meeting room Lions 1 (RDC)