Relay Policy Learning: Solving Long-Horizon Tasks via Imitation and Reinforcement Learning.
Abhishek GuptaVikash KumarCorey LynchSergey LevineKarol HausmanPublished in: CoRR (2019)
Keyphrases
- reinforcement learning
- learning algorithm
- learning process
- sequential decision making
- action selection
- transfer learning
- actor critic
- markov decision problems
- partially observable
- multiple tasks
- active learning
- learning tasks
- learning problems
- function approximation
- policy gradient
- dynamic programming
- model free reinforcement learning
- eligibility traces
- reinforcement learning agents
- imitation learning
- multi agent
- complex domains
- reinforcement learning algorithms
- multi task
- online learning
- supervised learning
- markov decision process
- state action
- rl algorithms
- machine learning
- solving problems
- optimal policy
- policy search
- state space
- solve complex tasks