PiCor: Multi-Task Deep Reinforcement Learning with Policy Correction.
Fengshuo BaiHongming ZhangTianyang TaoZhiheng WuYanna WangBo XuPublished in: AAAI (2023)
Keyphrases
- multi task
- reinforcement learning
- optimal policy
- transfer learning
- multi task learning
- policy search
- learning problems
- learning tasks
- function approximators
- multitask learning
- multiple tasks
- reward function
- markov decision processes
- feature selection
- multi class
- gaussian processes
- sparse learning
- state space
- learning algorithm
- labeled data
- learning models
- kernel methods
- supervised learning
- inductive learning
- active learning
- learning process
- training set
- machine learning