There Is No Turning Back: A Self-Supervised Approach for Reversibility-Aware Reinforcement Learning.
Nathan GrinsztajnJohan FerretOlivier PietquinPhilippe PreuxMatthieu GeistPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- function approximation
- reinforcement learning algorithms
- learning algorithm
- state space
- markov chain
- markov decision processes
- model free
- temporal difference learning
- optimal policy
- multi agent
- temporal difference
- machine learning
- direct policy search
- control problems
- action selection
- real time
- optimal control
- transfer learning
- monte carlo
- learning capabilities
- supervised learning
- markov decision process
- control policy
- learning agents
- information retrieval
- continuous state
- multi agent reinforcement learning