Off-Policy Evaluation in Partially Observable Environments.
Guy TennenholtzUri ShalitShie MannorPublished in: AAAI (2020)
Keyphrases
- partially observable environments
- policy evaluation
- partially observable markov decision processes
- reinforcement learning
- reinforcement learning algorithms
- temporal difference
- model free
- markov decision processes
- partially observable
- inverse reinforcement learning
- least squares
- monte carlo
- policy iteration
- function approximation
- finite state
- optimal policy
- belief state
- decision problems
- state space
- dynamical systems
- markov decision problems
- dynamic programming
- planning problems
- multi agent
- variance reduction
- action selection
- evaluation function
- reward function
- average cost
- supervised learning
- policy gradient