Uniform State Abstraction for Reinforcement Learning.
John BurdenDaniel KudenkoPublished in: ECAI (2020)
Keyphrases
- state abstraction
- reinforcement learning
- state space
- reinforcement learning agents
- markov decision processes
- hierarchical reinforcement learning
- reinforcement learning algorithms
- path finding
- function approximation
- model free
- initial state
- temporal difference
- learning algorithm
- optimal control
- supervised learning
- finite state
- transfer learning
- multi agent
- dynamic programming
- heuristic search
- sufficient conditions
- action space
- markov decision problems