The Tradeoff Between Generative and Discriminative Classifiers - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2004

The Tradeoff Between Generative and Discriminative Classifiers

Guillaume Bouchard
  • Function : Author
  • PersonId : 874970
Bill Triggs


Given any generative classifier based on an inexact density model, we can define a discriminative counterpart that reduces its asymptotic error rate. We introduce a family of classifiers that interpolate the two approaches, thus providing a new way to compare them and giving an estimation procedure whose classification performance is well balanced between the bias of generative classifiers and the variance of discriminative ones. We show that an intermediate trade-off between the two strategies is often preferable, both theoretically and in experiments on real data.
Fichier principal
Vignette du fichier
Bouchard-compstat04.pdf (106.42 Ko) Télécharger le fichier
Origin : Publisher files allowed on an open archive

Dates and versions

inria-00548546 , version 1 (20-12-2010)


  • HAL Id : inria-00548546 , version 1


Guillaume Bouchard, Bill Triggs. The Tradeoff Between Generative and Discriminative Classifiers. 16th IASC International Symposium on Computational Statistics (COMPSTAT '04), Aug 2004, Prague, Czech Republic. pp.721--728. ⟨inria-00548546⟩
902 View
1268 Download


Gmail Facebook Twitter LinkedIn More