Reinforcement learning for parameter estimation in statistical spoken dialogue systems.
Filip JurcícekBlaise ThomsonSteve J. YoungPublished in: Comput. Speech Lang. (2012)
Keyphrases
- parameter estimation
- spoken dialogue systems
- statistical models
- reinforcement learning
- maximum likelihood
- least squares
- markov random field
- dialogue management
- em algorithm
- model selection
- dialogue system
- human machine interaction
- random fields
- parameter estimation algorithm
- expectation maximization
- context aware
- model fitting
- multi domain
- estimation problems
- multi agent
- parameter estimates
- learning algorithm
- state space
- human users
- human computer interaction
- domain specific