On the Sample Efficiency of Abstractions and Potential-Based Reward Shaping in Reinforcement Learning.
Giuseppe CanonacoLeo ArdonAlberto PozancoDaniel BorrajoPublished in: CoRR (2024)
Keyphrases
- reward shaping
- reinforcement learning
- reinforcement learning algorithms
- complex domains
- learning algorithm
- optimal policy
- state space
- markov decision processes
- multi agent
- machine learning
- function approximation
- model free
- computational complexity
- transfer learning
- partially observable
- decision making
- markov decision problems