Exploration-Exploitation Trade-off in Reinforcement Learning on Online Markov Decision Processes with Global Concave Rewards.
Wang Chi CheungPublished in: CoRR (2019)
Keyphrases
- markov decision processes
- reinforcement learning
- optimal policy
- reinforcement learning algorithms
- state space
- finite state
- policy iteration
- reward function
- model based reinforcement learning
- dynamic programming
- decision theoretic planning
- partially observable
- action space
- reachability analysis
- planning under uncertainty
- markov decision process
- state and action spaces
- factored mdps
- sequential decision making under uncertainty
- state abstraction
- finite horizon
- average cost
- function approximation
- transition matrices
- infinite horizon
- objective function
- action sets
- average reward
- learning algorithm
- partially observable markov decision processes
- discounted reward
- total reward
- temporal difference
- approximate dynamic programming
- policy evaluation
- linear programming
- decentralized control
- continuous state spaces
- machine learning