Heuristic Approaches for Support Vector Machines with the Ramp Loss

Emilio Carrizosa, Amaya Nogales-Gómez, Dolores Romero Morales

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

Recently, Support Vector Machines with the ramp loss (RLM) have attracted attention from the computational point of view. In this technical note, we propose two heuristics, the first one based on solving the continuous relaxation of a Mixed Integer Nonlinear formulation of the RLM and the second one based on the training of an SVM classifier on a reduced dataset identified by an integer linear problem. Our computational results illustrate the ability of our heuristics to handle datasets of much larger size than those previously addressed in the literature
Recently, Support Vector Machines with the ramp loss (RLM) have attracted attention from the computational point of view. In this technical note, we propose two heuristics, the first one based on solving the continuous relaxation of a Mixed Integer Nonlinear formulation of the RLM and the second one based on the training of an SVM classifier on a reduced dataset identified by an integer linear problem. Our computational results illustrate the ability of our heuristics to handle datasets of much larger size than those previously addressed in the literature
LanguageEnglish
JournalOptimization Letters
Volume8
Issue number3
Pages1125-1135
ISSN1862-4472
DOIs
StatePublished - 2014
Externally publishedYes

Keywords

    Cite this

    Carrizosa, Emilio ; Nogales-Gómez, Amaya ; Morales, Dolores Romero. / Heuristic Approaches for Support Vector Machines with the Ramp Loss. In: Optimization Letters. 2014 ; Vol. 8, No. 3. pp. 1125-1135
    @article{06f3bdc4441a43598fb14e821e6a833e,
    title = "Heuristic Approaches for Support Vector Machines with the Ramp Loss",
    abstract = "Recently, Support Vector Machines with the ramp loss (RLM) have attracted attention from the computational point of view. In this technical note, we propose two heuristics, the first one based on solving the continuous relaxation of a Mixed Integer Nonlinear formulation of the RLM and the second one based on the training of an SVM classifier on a reduced dataset identified by an integer linear problem. Our computational results illustrate the ability of our heuristics to handle datasets of much larger size than those previously addressed in the literature",
    keywords = "Support vector machines, Ramp loss, Mixed integer nonlinear programming, Heuristics",
    author = "Emilio Carrizosa and Amaya Nogales-G{\'o}mez and Morales, {Dolores Romero}",
    year = "2014",
    doi = "10.1007/s11590-013-0630-9",
    language = "English",
    volume = "8",
    pages = "1125--1135",
    journal = "Optimization Letters",
    issn = "1862-4472",
    publisher = "Springer",
    number = "3",

    }

    Heuristic Approaches for Support Vector Machines with the Ramp Loss. / Carrizosa, Emilio; Nogales-Gómez, Amaya; Morales, Dolores Romero.

    In: Optimization Letters, Vol. 8, No. 3, 2014, p. 1125-1135.

    Research output: Contribution to journalJournal articleResearchpeer-review

    TY - JOUR

    T1 - Heuristic Approaches for Support Vector Machines with the Ramp Loss

    AU - Carrizosa,Emilio

    AU - Nogales-Gómez,Amaya

    AU - Morales,Dolores Romero

    PY - 2014

    Y1 - 2014

    N2 - Recently, Support Vector Machines with the ramp loss (RLM) have attracted attention from the computational point of view. In this technical note, we propose two heuristics, the first one based on solving the continuous relaxation of a Mixed Integer Nonlinear formulation of the RLM and the second one based on the training of an SVM classifier on a reduced dataset identified by an integer linear problem. Our computational results illustrate the ability of our heuristics to handle datasets of much larger size than those previously addressed in the literature

    AB - Recently, Support Vector Machines with the ramp loss (RLM) have attracted attention from the computational point of view. In this technical note, we propose two heuristics, the first one based on solving the continuous relaxation of a Mixed Integer Nonlinear formulation of the RLM and the second one based on the training of an SVM classifier on a reduced dataset identified by an integer linear problem. Our computational results illustrate the ability of our heuristics to handle datasets of much larger size than those previously addressed in the literature

    KW - Support vector machines

    KW - Ramp loss

    KW - Mixed integer nonlinear programming

    KW - Heuristics

    U2 - 10.1007/s11590-013-0630-9

    DO - 10.1007/s11590-013-0630-9

    M3 - Journal article

    VL - 8

    SP - 1125

    EP - 1135

    JO - Optimization Letters

    T2 - Optimization Letters

    JF - Optimization Letters

    SN - 1862-4472

    IS - 3

    ER -