Optimal linear regression on classifier outputs

Abstract : We consider the combination of the outputs of several classifiers trained independently for the same discrimination task. We introduce new results which provide optimal solutions in the case of linear combinations. We compare our solutions to existing ensemble methods and characterize situations where our approach should be preferred.
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01649040
Contributor : Lip6 Publications <>
Submitted on : Monday, November 27, 2017 - 10:39:24 AM
Last modification on : Thursday, March 21, 2019 - 1:06:50 PM

Links full text

Identifiers

Citation

Yann Guermeur, Florence d'Alché-Buc, Patrick Gallinari. Optimal linear regression on classifier outputs. ICANN'97 - 7th International Conference of Artificial Neural Networks, Oct 1997, Lausanne, Switzerland. pp.481-486, ⟨10.1007/BFb0020201⟩. ⟨hal-01649040⟩

Share

Metrics

Record views

62