Adaptive Reward Shifting Based on Behavior Proximity for Offline Reinforcement Learning.
Zhe ZhangXiaoyang TanPublished in: IJCAI (2023)
Keyphrases
- reinforcement learning
- adaptive behavior
- adaptive control
- multi agent environments
- learning algorithm
- learning capabilities
- reinforcement learning algorithms
- function approximation
- eligibility traces
- state space
- optimal policy
- model free
- machine learning
- temporal difference
- multi agent
- markov decision processes
- behavior patterns
- partially observable environments
- dynamic programming
- real robot
- real time
- action selection
- genetic algorithm
- partially observable
- human behavior
- reward function
- optimal control
- learning problems
- transfer learning
- mobile robot
- learning process
- neural network