Hoeffding and Bernstein races for selecting policies in evolutionary direct policy search.

Verena Heidrich-Meisner Christian Igel

Published in: ICML (2009)

Keyphrases

direct policy search
reinforcement learning
optimal policy
genetic algorithm
multi agent
evolutionary computation
state space
expert systems
support vector machine
supervised learning
radial basis function
function approximation
partially observable markov decision processes