Action Advising with Advice Imitation in Deep Reinforcement Learning.
Ercument IlhanJeremy GowDiego Perez LiebanaPublished in: AAMAS (2021)
Keyphrases
- reinforcement learning
- action selection
- action space
- reward shaping
- state action
- partially observable domains
- state space
- function approximation
- multi agent
- markov decision processes
- continuous state
- temporal difference
- reinforcement learning algorithms
- transition model
- learning algorithm
- optimal policy
- agent learns
- fitted q iteration
- deep learning
- model free
- dynamic programming
- machine learning
- function approximators
- temporal difference learning
- learning classifier systems
- imitation learning
- human actions
- genetic algorithm
- neural network