Dialogue manager domain adaptation using Gaussian process reinforcement learning.
Milica GasicNikola MrksicLina Maria Rojas-BarahonaPei-Hao SuStefan UltesDavid VandykeTsung-Hsien WenSteve J. YoungPublished in: Comput. Speech Lang. (2017)
Keyphrases
- gaussian process
- domain adaptation
- reinforcement learning
- semi supervised
- transfer learning
- labeled data
- co training
- semi supervised learning
- bayesian framework
- cross domain
- regression model
- model selection
- multiple sources
- latent variables
- active learning
- hyperparameters
- unlabeled data
- target domain
- learning algorithm
- machine learning
- generative model
- state space
- unsupervised learning
- test data
- multi view
- supervised learning
- dialogue system
- machine learning algorithms
- pairwise
- markov decision processes
- knowledge transfer
- optimal policy
- learning process
- training data