Title | ||
---|---|---|
Reinforcement Learning For Dialog Management Using Least-Squares Policy Iteration And Fast Feature Selection |
Abstract | ||
---|---|---|
Reinforcement learning (RL) is a promising technique for creating a dialog manager. RL accepts features of the current dialog state and seeks to find the best action given those features. Although it is often easy to posit a large set of potentially useful features, in practice, it is difficult to find the subset which is large enough to contain useful information yet compact enough to reliably learn a good policy. In this paper, we propose a method for RL optimization which automatically performs feature selection. The algorithm is based on least-squares policy iteration. a state-of-the-art RL algorithm which is highly sample-efficient and can learn from a static corpus or on-line. Experiments in dialog simulation show it is more stable than a baseline RL algorithm taken from a working dialog system. |
Year | Venue | Keywords |
---|---|---|
2009 | INTERSPEECH 2009: 10TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2009, VOLS 1-5 | Dialog management, spoken dialog systems, partially observable Markov decision processes |
Field | DocType | Citations |
Least squares,Dialog box,Feature selection,Pattern recognition,Computer science,Artificial intelligence,Dialog management,Dialog system,Machine learning,Reinforcement learning | Conference | 25 |
PageRank | References | Authors |
1.05 | 12 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Lihong Li | 1 | 2390 | 128.53 |
Jason D. Williams | 2 | 1319 | 76.49 |
Suhrid Balakrishnan | 3 | 238 | 14.60 |