TIRL: Enriching Actor-Critic RL with non-expert human teachers and a Trust Model.
Félix RutardOlivier SigaudMohamed ChetouaniPublished in: RO-MAN (2020)
Keyphrases
- trust model
- actor critic
- reinforcement learning
- policy gradient
- temporal difference
- reinforcement learning algorithms
- gradient method
- approximate dynamic programming
- optimal control
- neuro fuzzy
- policy iteration
- function approximation
- learning process
- multiagent systems
- model free
- state space
- multi agent systems
- transfer learning
- machine learning
- average reward
- cooperative
- learning algorithm
- monte carlo
- markov decision processes
- multi agent
- e learning