Feudal Reinforcement Learning for Dialogue Management in Large Domains.
Iñigo CasanuevaPawel BudzianowskiPei-Hao SuStefan UltesLina Maria Rojas-BarahonaBo-Hsiang TsengMilica GasicPublished in: NAACL-HLT (2) (2018)
Keyphrases
- dialogue management
- reinforcement learning
- dialogue system
- learning agent
- spoken dialogue systems
- transfer learning
- function approximation
- natural language generation
- state space
- markov decision processes
- language understanding
- machine learning
- optimal control
- learning process
- model free
- dynamic programming
- multi agent
- partially observable markov decision process
- machine translation
- learning algorithm