Distributed Structured Actor-Critic Reinforcement Learning for Universal Dialogue Management.
Zhi ChenLu ChenXiaoyuan LiuKai YuPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2020)
Keyphrases
- actor critic
- dialogue management
- reinforcement learning
- temporal difference
- dialogue system
- policy gradient
- optimal control
- approximate dynamic programming
- learning agent
- reinforcement learning algorithms
- function approximation
- gradient method
- neuro fuzzy
- multi agent
- policy iteration
- spoken dialogue systems
- markov decision processes
- machine learning
- average reward
- partially observable markov decision processes
- learning algorithm
- language understanding
- optimal policy
- natural language generation
- state space
- markov chain
- transfer learning