PAC-Bayesian Policy Evaluation for Reinforcement Learning
Mahdi Milani FardJoelle PineauCsaba SzepesváriPublished in: CoRR (2012)
Keyphrases
- policy evaluation
- pac bayesian
- reinforcement learning
- temporal difference
- least squares
- model free
- distribution free
- function approximation
- monte carlo
- policy iteration
- markov decision processes
- reinforcement learning algorithms
- variance reduction
- optimal policy
- state space
- semi parametric
- dynamic programming
- generalization bounds
- learning algorithm
- machine learning
- partially observable markov decision processes
- learning problems
- sufficient conditions
- optimal control
- evaluation function
- supervised learning
- probabilistic model