Switch-based Active Deep Dyna-Q: Efficient Adaptive Planning for Task-Completion Dialogue Policy Learning.
Yuexin WuXiujun LiJingjing LiuJianfeng GaoYiming YangPublished in: CoRR (2018)
Keyphrases
- learning algorithm
- learning process
- mixed initiative
- knowledge acquisition
- action selection
- adaptive learning
- learning tasks
- knowledge level
- adaptive control
- learning systems
- stochastic domains
- neural network
- function approximators
- domain independent
- optimal policy
- unsupervised learning
- supervised learning
- reinforcement learning
- bayesian networks
- machine learning