Interpretable Reward Redistribution in Reinforcement Learning: A Causal Approach.
Yudi ZhangYali DuBiwei HuangZiyan WangJun WangMeng FangMykola PechenizkiyPublished in: NeurIPS (2023)
Keyphrases
- reinforcement learning
- function approximation
- state space
- model free
- reinforcement learning algorithms
- learning algorithm
- bayesian networks
- eligibility traces
- markov decision processes
- reward function
- causal networks
- partially observable environments
- causal models
- multi agent
- temporal difference
- total reward
- average reward
- machine learning
- causal relationships
- action selection
- learning problems
- dynamic programming
- reward shaping
- neural network
- policy iteration
- learning agent
- action space
- reinforcement learning methods
- policy gradient
- optimal control
- policy search
- causal theories
- causal interactions