Material for the Sequential Decision Making Reading Group

Here are some suggested background readings on Reinforcement Learning and Multiagent Reinforcement Learning.

Themes	Suggested Readings (by)
1. Bayesian RL / Bayesian exploration / Bandit problems	Georgios
2. Bayesian networks	Georgios
3. Factored MDPs and related solution methods	Georgios
4. DecMDPs / DecPOMDPs
5. First-order MDPs
6. Gaussian Processes (GPs) / GPs +RL / GPs+ Machine Learning
7. Inverse RL	Georgios
8. Coordination Games / Learning in Coordination Games
9. Hierarchical RL and Abstraction methods	Georgios
10.Hidden Markov Models
11.Uncertainty and Learning in Games

Schedule

Date	Title	Discussion Leaders (you can find links to slides here, if available; check links at leaders’ names)
24 Feb. 2009	Sutton and Barto, Ch. 1, 2, 3, 4, 5, 6	Simon, Enrique, Georgios
4 Mar. 2009	Sutton and Barto, Ch. 6 (cont.) ; Michael Duff’s thesis, Ch. 1 (focus on Bayes-Adaptive MDPs / optimal learning formulation)	Simon, Enrique, Georgios
25 Mar. 2009	1) Bayesian Q-Learning (http://www.cs.huji.ac.il/~nir/Abstracts/DFR1.html) 2) Model-Based Bayesian Exploration (http://robotics.stanford.edu/users/nir/Papers/DFA1.pdf) 3) A Bayesian Framework for Reinforcement Learning (http://citeseer.ist.psu.edu/old/522507.html) People should find the following videotalk on “Model-Based Bayesian RL” by Pascal Poupart very informative (slides can also be downloaded from the site): http://videolectures.net/icml07_poupart_mbbrl/	1) Long 2) Archie 3) Kate
7 Apr. 2009	Model-based Bayesian Reinforcement Learning in Partially Observable Domains by Pascal Poupart and Nikos Vlassis. AI-Math 2008	- Simon (George will also be speaking briefly about the following paper: Poupart, Vlassis, Hoey and Regan: An analytic solution to Discrete Bayesian Reinforcement Learning, 23^rd ICML, 2006 )
28 Apr. 2009	Introduction to Bayesian Networks (D.Heckerman: A tutorial on Bayesian Networks: Sections 1-5)	- Kate, Long
27 May 2009	More on Bayesian Networks: Sections 6-12	- Long, Simon
1 Jul 2009	“Optimizing Information Exchange in Cooperative Multiagent Systems” by Claudia Goldman and Shlomo Zilberstein	- George, Zinovi
29 Jul 2009	On the Value of Communication during Online Multiagent Planning under Uncertainty	- Simon [POSTPONED]
21 Oct 2009	DEBATE: Bayesian vs. Frequentist - http://www-stat.stanford.edu/~ckirby/brad/papers/2005NEWModernScience.pdf - http://nb.vse.cz/kfil/elogos/science/vallverdu08.pdf	- Zinovi, Enrique, George et al.

Page maintained by Georgios Chalkiadakis.