Bayesian nonparametric models for ranked data

Abstract : We develop a Bayesian nonparametric extension of the popular Plackett-Luce choice model that can handle an infinite number of choice items. Our framework is based on the theory of random atomic measures, with the prior specified by a gamma process. We derive a posterior characterization and a simple and effective Gibbs sampler for posterior simulation. We develop a time-varying extension of our model, and apply it to the New York Times lists of weekly bestselling books.
Document type :
Conference papers
NIPS - Neural Information Processing Systems, Dec 2012, Lake Tahoe, United States. MIT Press, 2012


https://hal.inria.fr/hal-00730792
Contributor : Francois Caron <>
Submitted on : Sunday, November 18, 2012 - 8:15:29 PM
Last modification on : Monday, November 19, 2012 - 8:40:54 AM

Files

RR-8140.pdf
fileSource_public_author

Identifiers

  • HAL Id : hal-00730792, version 1
  • ARXIV : 1211.4321

Collections

Citation

Francois Caron, Yee Whye Teh. Bayesian nonparametric models for ranked data. NIPS - Neural Information Processing Systems, Dec 2012, Lake Tahoe, United States. MIT Press, 2012. <hal-00730792>

Export

Share

Metrics

Consultation de
la notice

179

Téléchargement du document

33