High-Confidence Policy Optimization: Reshaping Ambiguity Sets in Robust MDPs.

Bahram Behzadian Reazul Hasan Russel Marek Petrik

Published in: CoRR (2019)

Keyphrases

high confidence
optimal policy
markov decision processes
markov decision process
association rules
reinforcement learning
data sets
markov decision problems
finite horizon
policy iteration
robust optimization
data mining
policy search
finite state
text classification
prior knowledge
pairwise
factored mdps