Reinforcement Learning in Latent Action Sequence Space.
Heecheol KimMasanori YamadaKosuke MiyoshiTomoharu IwataHiroshi YamakawaPublished in: IROS (2020)
Keyphrases
- reinforcement learning
- action space
- action selection
- function approximation
- multi agent
- partially observable domains
- optimal policy
- search space
- state space
- hidden state
- higher dimensional
- vector space
- dimensional vector
- reward shaping
- action sequences
- temporal difference
- latent variables
- data sets
- input data
- supervised learning
- data points
- dynamic programming
- learning algorithm