Computer Science – Artificial Intelligence
Scientific paper
2010-02-07
Computer Science
Artificial Intelligence
8 pages, 3 figures, 3 tables
Scientific paper
Adaptive control problems are notoriously difficult to solve even in the presence of plant-specific controllers. One way to by-pass the intractable computation of the optimal policy is to restate the adaptive control as the minimization of the relative entropy of a controller that ignores the true plant dynamics from an informed controller. The solution is given by the Bayesian control rule-a set of equations characterizing a stochastic adaptive controller for the class of possible plant dynamics. Here, the Bayesian control rule is applied to derive BCR-MDP, a controller to solve undiscounted Markov decision processes with finite state and action spaces and unknown dynamics. In particular, we derive a non-parametric conjugate prior distribution over the policy space that encapsulates the agent's whole relevant history and we present a Gibbs sampler to draw random policies from this distribution. Preliminary results show that BCR-MDP successfully avoids sub-optimal limit cycles due to its built-in mechanism to balance exploration versus exploitation.
Braun Daniel A.
Ortega Pedro A.
No associations
LandOfFree
A Minimum Relative Entropy Controller for Undiscounted Markov Decision Processes does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with A Minimum Relative Entropy Controller for Undiscounted Markov Decision Processes, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and A Minimum Relative Entropy Controller for Undiscounted Markov Decision Processes will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-582847