Transfer Learning based Task-oriented Dialogue Policy for Multiple Domains using Hierarchical Reinforcement Learning.
Tulika SahaSriparna SahaPushpak BhattacharyyaPublished in: IJCNN (2020)
Keyphrases
- transfer learning
- multiple domains
- hierarchical reinforcement learning
- cross domain
- reinforcement learning
- optimal policy
- reward function
- markov decision process
- average reward
- knowledge transfer
- learning tasks
- target domain
- model free
- state abstraction
- labeled data
- policy iteration
- transfer knowledge
- machine learning
- state space
- markov decision processes
- active learning
- partially observable markov decision processes
- semi supervised learning
- function approximation
- text categorization
- text classification
- reinforcement learning algorithms
- collaborative filtering
- temporal difference
- dynamic programming
- learning process
- learning algorithm
- infinite horizon
- unlabeled data
- unsupervised learning
- natural language
- training data
- neural network