Composite Task-Completion Dialogue Policy Learning via Hierarchical Deep Reinforcement Learning.
Baolin PengXiujun LiLihong LiJianfeng GaoAsli CelikyilmazSungjin LeeKam-Fai WongPublished in: EMNLP (2017)
Keyphrases
- reinforcement learning
- learning process
- learning algorithm
- supervised learning
- learning tasks
- policy search
- online learning
- unsupervised learning
- learning systems
- function approximation
- learning agents
- partially observable environments
- neural network
- hierarchical reinforcement learning
- policy gradient
- deep learning
- partially observable
- learning capabilities
- reinforcement learning algorithms
- model free
- optimal policy
- human computer
- state action
- continuous state
- markov decision processes
- eligibility traces