A Game-Theoretic Perspective of Generalization in Reinforcement Learning.
Chang YangRuiyu WangXinrun WangZhen WangPublished in: CoRR (2022)
Keyphrases
- game theoretic perspective
- reinforcement learning
- function approximation
- machine learning
- action selection
- state space
- policy search
- dynamic programming
- reinforcement learning algorithms
- transfer learning
- markov decision processes
- databases
- relational reinforcement learning
- learning agents
- partially observable
- temporal difference
- learning algorithm
- knowledge base
- optimal policy
- dynamic environments
- control system
- learning environment
- multi agent
- information systems