Reward-Balancing for Statistical Spoken Dialogue Systems using Multi-objective Reinforcement Learning.
Stefan UltesPawel BudzianowskiIñigo CasanuevaNikola MrksicLina Maria Rojas-BarahonaPei-Hao SuTsung-Hsien WenMilica GasicSteve J. YoungPublished in: SIGDIAL Conference (2017)
Keyphrases
- reinforcement learning
- multi objective
- spoken dialogue systems
- dialogue management
- dialogue system
- evolutionary algorithm
- multi objective optimization
- human machine interaction
- multi domain
- state space
- objective function
- reinforcement learning algorithms
- eligibility traces
- context aware
- model free
- learning algorithm
- machine learning
- genetic algorithm
- markov decision processes
- temporal difference
- partially observable
- learning agent
- multi agent
- optimal policy
- domain specific
- general purpose
- supervised learning