DynaSTI: Dynamics modeling with sequential temporal information for reinforcement learning in Atari.
Jaehoon KimYoung Jae LeeMingu KwakYoungjoon ParkSeoung Bum KimPublished in: Knowl. Based Syst. (2024)
Keyphrases
- temporal information
- reinforcement learning
- temporal evolution
- temporal aspects
- video sequences
- temporal data
- temporal reasoning
- temporal patterns
- temporal dimension
- spatial information
- contextual information
- temporal constraints
- linear value function approximation
- temporal context
- temporal expressions
- temporal knowledge
- model free
- function approximation
- machine learning
- markov decision processes
- supervised learning
- multi agent
- computer vision