The Principle of Unchanged Optimality in Reinforcement Learning Generalization.
Alex IrpanXingyou SongPublished in: CoRR (2019)
Keyphrases
- reinforcement learning
- markov decision processes
- model free
- optimal solution
- dynamic programming
- function approximation
- state space
- reinforcement learning algorithms
- temporal difference
- optimal policy
- data sets
- learning process
- learning algorithm
- machine learning
- supervised learning
- multi agent
- genetic algorithm
- database
- average cost
- autonomous learning