Timezone: »

Active Preference Learning with Discrete Choice Data
Eric Brochu · Nando de Freitas · Abhijeet Ghosh

Mon Dec 03 10:30 AM -- 10:40 AM (PST) @

We propose an active learning algorithm that learns a continuous valuation model from discrete preferences. The algorithm automatically decides what items are best presented to an individual in order to find the item that they value highly in as few trials as possible, and exploits quirks of human psychology to minimize time and cognitive burden. To do this, our algorithm maximizes the expected improvement at each query without accurately modelling the entire valuation surface, which would be needlessly expensive. The problem is particularly difficult because the space of choices is infinite. We demonstrate the effectiveness of the new algorithm compared to related active learning methods. We also embed the algorithm within a decision making tool for assisting digital artists in rendering materials. The tool finds the best parameters while minimizing the number of queries.

Author Information

Eric Brochu (University of British Columbia)
Nando de Freitas (University of Oxford)
Abhijeet Ghosh

More from the Same Authors