Learning Markov State Abstractions for Deep Reinforcement Learning.
Cameron AllenNeev ParikhOmer GottesmanGeorge KonidarisPublished in: NeurIPS (2021)
Keyphrases
- reinforcement learning
- learning process
- learning systems
- learning problems
- learning algorithm
- learning tasks
- active learning
- state abstraction
- deep architectures
- supervised learning
- markov chain
- probabilistic model
- dynamic programming
- partially observable
- admissible heuristics
- learning agents
- hidden state
- autonomous learning