Leveraging Reward Consistency for Interpretable Feature Discovery in Reinforcement Learning.
Qisen YangHuanqian WangMukun TongWenjie ShiGao HuangShiji SongPublished in: IEEE Trans. Syst. Man Cybern. Syst. (2024)
Keyphrases
- reinforcement learning
- function approximation
- eligibility traces
- markov decision processes
- reward function
- multi agent
- image features
- consistency checking
- machine learning
- knowledge discovery
- state space
- constraint networks
- function approximators
- reinforcement learning algorithms
- data mining
- optimal control
- feature vectors
- temporal difference
- action selection
- path consistency
- reinforcement learning methods
- partially observable environments