A new discriminative kernel from probabilistic models

K. Tsuda, M. Kawanabe, G. Ratsch, S. Sonnenburg, Klaus Muller

Research output: Chapter in Book/Report/Conference proceedingConference contribution

12 Citations (Scopus)

Abstract

Recently. Jaakkola and Haussler proposed a method for constructing kernel functions from probabilistic models. Their so called "Fisher kernel" has been combined with discriminative classifiers such as SVM and applied successfully in e.g. DNA and protein analysis. Whereas the Fisher kernel (FK) is calculated from the marginal log-likelihood, we propose the TOP kernel derived from Tangent vectors Of Posterior log-odds. Furthermore we develop a theoretical framework on feature extractors from probabilistic models and use it for analyzing FK and TOP. In experiments our new discriminative TOP kernel compares favorably to the Fisher kernel.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems
PublisherNeural information processing systems foundation
ISBN (Print)0262042088, 9780262042086
Publication statusPublished - 2002 Jan 1
Externally publishedYes
Event15th Annual Neural Information Processing Systems Conference, NIPS 2001 - Vancouver, BC, Canada
Duration: 2001 Dec 32001 Dec 8

Other

Other15th Annual Neural Information Processing Systems Conference, NIPS 2001
CountryCanada
CityVancouver, BC
Period01/12/301/12/8

    Fingerprint

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Tsuda, K., Kawanabe, M., Ratsch, G., Sonnenburg, S., & Muller, K. (2002). A new discriminative kernel from probabilistic models. In Advances in Neural Information Processing Systems Neural information processing systems foundation.