O. Mangasarian. Advances in Large Margin Classifiers, page 135-146. Cambridge, MA, MIT Press, (2000)
Abstract
Arbitrary kernel functions which need not satisfy Mercer's condition
can be used. This goal is achieved by separating the regularizer
from the actual separation condition. For quadratic regularization
this leads to a convex quadratic program that is no more difficult
to solve than the standard SV optimization problem. Sparse expansions
are achieved when the $1$-norm of the expansion coefficients is
chosen to restrict the class of admissible functions. The problems
are formulated in a way which is compatible with Mathematical Programming
literature.
%0 Conference Paper
%1 Mangasarian00
%A Mangasarian, O.
%B Advances in Large Margin Classifiers
%C Cambridge, MA
%D 2000
%E Smola, A.J.
%E Bartlett, P.L.
%E Schölkopf, B.
%E Schuurmans, D.
%I MIT Press
%K imported
%P 135-146
%T Generalized Support Vector Machines
%X Arbitrary kernel functions which need not satisfy Mercer's condition
can be used. This goal is achieved by separating the regularizer
from the actual separation condition. For quadratic regularization
this leads to a convex quadratic program that is no more difficult
to solve than the standard SV optimization problem. Sparse expansions
are achieved when the $1$-norm of the expansion coefficients is
chosen to restrict the class of admissible functions. The problems
are formulated in a way which is compatible with Mathematical Programming
literature.
@inproceedings{Mangasarian00,
abstract = {Arbitrary kernel functions which need not satisfy Mercer's condition
can be used. This goal is achieved by separating the regularizer
from the actual separation condition. For quadratic regularization
this leads to a convex quadratic program that is no more difficult
to solve than the standard SV optimization problem. Sparse expansions
are achieved when the $1$-norm of the expansion coefficients is
chosen to restrict the class of admissible functions. The problems
are formulated in a way which is compatible with Mathematical Programming
literature.},
added-at = {2008-04-30T12:59:47.000+0200},
address = {Cambridge, MA},
author = {Mangasarian, O.},
biburl = {https://www.bibsonomy.org/bibtex/259ef4fd60a9948cdc33f104f4f46b1b7/kdubiq},
booktitle = {Advances in Large Margin Classifiers},
description = {KDubiq Blueprint},
editor = {Smola, A.J. and Bartlett, P.L. and Sch{\"o}lkopf, B. and Schuurmans, D.},
groupsearch = {0},
interhash = {5f6d81f1a3dacf272567e28eeec7ff19},
intrahash = {59ef4fd60a9948cdc33f104f4f46b1b7},
keywords = {imported},
pages = {135-146},
publisher = {{MIT} Press},
timestamp = {2008-04-30T13:00:23.000+0200},
title = {Generalized Support Vector Machines},
year = 2000
}