Discrete Factorial Representations as an Abstraction for Goal Conditioned Reinforcement Learning.
Riashat IslamHongyu ZangAnirudh GoyalAlex LambKenji KawaguchiXin LiRomain LarocheYoshua BengioRemi Tachet des CombesPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- high level
- continuous state and action spaces
- learning algorithm
- state abstraction
- data mining
- machine learning
- hidden markov models
- discrete version
- continuous state
- independent component analysis
- state space
- data sets
- signal processing
- learning problems
- finite number
- learning process
- genetic algorithm
- neural network