bluelogo.gif (1643 bytes)

home page

Using Free Energies to Represent Q-values in a Multiagent Reinforcement Learning Task

Brian Sallans
Department of Computer Science
University of Toronto
Toronto M5S 2Z9 Canada

Geoffrey Hinton
Gatsby Computational Neuroscience Unit
University College London
17 Queen Square, London WC1N 3AR, UK


The problem of reinforcement learning in large factored Markov decision processes is explored.  The Q-value of a state-action pair is approximated by the free energy of a product of experts network.  Network parameters are learned on-line using a modified SARSA algorithm which minimizes the inconsistency of the Q-values of consecutive state-action pairs.  Actions are chosen based on the current value estimates by fixing the current state and sampling actions from the network using Gibbs sampling.  The algorithm is tested on a co-operative multi-agent task.  The product of experts model is found to perform comparably to table-based Q-learning for small instances of the task, and continues to perform well when the problem becomes too large for a table-based representation.

 Download  [pdf] [ps.gz]

Submitted to Advances in Neural Information Processing Systems 13, MIT Press, Cambridge, MA

[home page]  [publications]