Contrastive Visual Explanations for Reinforcement Learning via Counterfactual Rewards.
Xiaowei LiuKevin McAreaveyWeiru LiuPublished in: xAI (2) (2023)
Keyphrases
- reinforcement learning
- markov decision processes
- state space
- function approximation
- reinforcement learning algorithms
- model free
- multi agent
- visual features
- visual representations
- temporal difference
- visual information
- machine learning
- reward shaping
- action selection
- visual perception
- low level
- image retrieval
- optimal control
- learning problems
- transfer learning
- dynamic programming
- partially observable
- visual representation
- markov decision process
- transition model
- bandit problems
- high level