A Scatter-Based Prototype Framework and Multi-Class Extension of Support Vector Machines

Robert Jenssen, Marius Kloft, Alexander Zien, Sören Sonnenburg, Klaus Muller

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

We provide a novel interpretation of the dual of support vector machines (SVMs) in terms of scatter with respect to class prototypes and their mean. As a key contribution, we extend this framework to multiple classes, providing a new joint Scatter SVM algorithm, at the level of its binary counterpart in the number of optimization variables. This enables us to implement computationally efficient solvers based on sequential minimal and chunking optimization. As a further contribution, the primal problem formulation is developed in terms of regularized risk minimization and the hinge loss, revealing the score function to be used in the actual classification of test patterns. We investigate Scatter SVM properties related to generalization ability, computational efficiency, sparsity and sensitivity maps, and report promising results.

Original languageEnglish
Article numbere42947
JournalPLoS One
Volume7
Issue number10
DOIs
Publication statusPublished - 2012 Oct 30

Fingerprint

prototypes
Support vector machines
Hinges
Computational efficiency
Research Design
Joints
support vector machines
Support Vector Machine
testing

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Medicine(all)

Cite this

A Scatter-Based Prototype Framework and Multi-Class Extension of Support Vector Machines. / Jenssen, Robert; Kloft, Marius; Zien, Alexander; Sonnenburg, Sören; Muller, Klaus.

In: PLoS One, Vol. 7, No. 10, e42947, 30.10.2012.

Research output: Contribution to journalArticle

Jenssen, Robert ; Kloft, Marius ; Zien, Alexander ; Sonnenburg, Sören ; Muller, Klaus. / A Scatter-Based Prototype Framework and Multi-Class Extension of Support Vector Machines. In: PLoS One. 2012 ; Vol. 7, No. 10.
@article{3a2adb0f0c2345a5937cd3e96bae2370,
title = "A Scatter-Based Prototype Framework and Multi-Class Extension of Support Vector Machines",
abstract = "We provide a novel interpretation of the dual of support vector machines (SVMs) in terms of scatter with respect to class prototypes and their mean. As a key contribution, we extend this framework to multiple classes, providing a new joint Scatter SVM algorithm, at the level of its binary counterpart in the number of optimization variables. This enables us to implement computationally efficient solvers based on sequential minimal and chunking optimization. As a further contribution, the primal problem formulation is developed in terms of regularized risk minimization and the hinge loss, revealing the score function to be used in the actual classification of test patterns. We investigate Scatter SVM properties related to generalization ability, computational efficiency, sparsity and sensitivity maps, and report promising results.",
author = "Robert Jenssen and Marius Kloft and Alexander Zien and S{\"o}ren Sonnenburg and Klaus Muller",
year = "2012",
month = "10",
day = "30",
doi = "10.1371/journal.pone.0042947",
language = "English",
volume = "7",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "10",

}

TY - JOUR

T1 - A Scatter-Based Prototype Framework and Multi-Class Extension of Support Vector Machines

AU - Jenssen, Robert

AU - Kloft, Marius

AU - Zien, Alexander

AU - Sonnenburg, Sören

AU - Muller, Klaus

PY - 2012/10/30

Y1 - 2012/10/30

N2 - We provide a novel interpretation of the dual of support vector machines (SVMs) in terms of scatter with respect to class prototypes and their mean. As a key contribution, we extend this framework to multiple classes, providing a new joint Scatter SVM algorithm, at the level of its binary counterpart in the number of optimization variables. This enables us to implement computationally efficient solvers based on sequential minimal and chunking optimization. As a further contribution, the primal problem formulation is developed in terms of regularized risk minimization and the hinge loss, revealing the score function to be used in the actual classification of test patterns. We investigate Scatter SVM properties related to generalization ability, computational efficiency, sparsity and sensitivity maps, and report promising results.

AB - We provide a novel interpretation of the dual of support vector machines (SVMs) in terms of scatter with respect to class prototypes and their mean. As a key contribution, we extend this framework to multiple classes, providing a new joint Scatter SVM algorithm, at the level of its binary counterpart in the number of optimization variables. This enables us to implement computationally efficient solvers based on sequential minimal and chunking optimization. As a further contribution, the primal problem formulation is developed in terms of regularized risk minimization and the hinge loss, revealing the score function to be used in the actual classification of test patterns. We investigate Scatter SVM properties related to generalization ability, computational efficiency, sparsity and sensitivity maps, and report promising results.

UR - http://www.scopus.com/inward/record.url?scp=84868100924&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84868100924&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0042947

DO - 10.1371/journal.pone.0042947

M3 - Article

VL - 7

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 10

M1 - e42947

ER -