A Maximum-Entropy Approach to Off-Policy Evaluation in Average-Reward MDPs.
Nevena LazicDong YinMehrdad FarajtabarNir LevineDilan GörürChris HarrisDale SchuurmansPublished in: NeurIPS (2020)
Keyphrases
- maximum entropy
- average reward
- policy evaluation
- markov decision processes
- policy iteration
- model free
- optimal policy
- reinforcement learning
- finite state
- state space
- fixed point
- markov models
- temporal difference
- dynamic programming
- least squares
- conditional random fields
- reinforcement learning algorithms
- markov decision process
- function approximation
- partially observable markov decision processes
- markov decision problems
- average cost
- machine learning
- infinite horizon
- state action
- policy gradient
- image segmentation
- learning algorithm
- reward function
- decision theoretic
- optimal control
- monte carlo
- heuristic search