ICML / COLT / UAI Review Session

The goal of the review session is to provide a flavor / highlights of
the conferences by presenting short summaries of some of the papers
presented at these conferences.  The following people have been will
be discussing the following papers:

- Bryan Singer
  - Lagoudakis and Littman's paper on Algorithm Selection using RL
- John Langford
  - Andrew Ng's paper on Inverse Reinforcement Learning
- Daniel Nikovski
  - Andrew Ng and Mike Jordan's paper on PEGASUS, their
    system for planning in POMDPs
- Sean Slattery
  - Goldman and Zhou's paper on Enhancing Supervised Learning 
    with Unlabeled Data
- Chuck Rosenberg
  - Bay and Pazzani's paper on Characterizing Model Errors 
    and Differences