Reinforcement Learning in Reward-Mixing MDPs.
Jeongyeol KwonYonathan EfroniConstantine CaramanisShie MannorPublished in: NeurIPS (2021)
Keyphrases
- reinforcement learning
- markov decision processes
- reward function
- average reward
- state space
- optimal policy
- function approximation
- model free
- learning algorithm
- machine learning
- eligibility traces
- reinforcement learning algorithms
- markov decision process
- temporal difference
- markov decision problems
- state and action spaces
- continuous state and action spaces
- partially observable
- action sets
- decision theoretic planning
- multi agent
- control problems
- reward shaping
- discounted reward
- dynamic programming
- continuous state
- rl algorithms
- action space
- policy iteration
- policy evaluation
- policy search
- optimal control
- learning tasks
- reinforcement learning methods
- long run