Universum parametric-margin ν-support vector machine for classification using the difference of convex functions algorithm

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F21%3A10451766" target="_blank" >RIV/00216208:11320/21:10451766 - isvavai.cz</a>

Výsledek na webu

<a href="https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=rti~2WpfKr" target="_blank" >https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=rti~2WpfKr</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/s10489-021-02402-6" target="_blank" >10.1007/s10489-021-02402-6</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Universum parametric-margin ν-support vector machine for classification using the difference of convex functions algorithm

Popis výsledku v původním jazyce

Universum data that do not belong to any class of a classification problem can be exploited to utilize prior knowledge to improve generalization performance. In this paper, we design a novel parametric ν-support vector machine with universum data (UPar-ν-SVM). Unlabeled samples can be integrated into supervised learning by means of UPar-ν-SVM. We propose a fast method to solve the suggested problem of UPar-ν-SVM. The primal problem of UPar-ν-SVM, which is a nonconvex optimization problem, is transformed into an unconstrained optimization problem so that the objective function can be treated as a difference of two convex functions (DC). To solve this unconstrained problem, a boosted difference of convex functions algorithm (BDCA) based on a generalized Newton method is suggested (named DC-UPar-ν-SVM). We examined our approach on UCI benchmark data sets, NDC data sets, a handwritten digit recognition data set, and a landmine detection data set. The experimental results confirmed the effectiveness and superiority of the proposed method for solving classification problems in comparison with other methods. (C) 2021, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.

Název v anglickém jazyce

Universum parametric-margin ν-support vector machine for classification using the difference of convex functions algorithm

Popis výsledku anglicky

Universum data that do not belong to any class of a classification problem can be exploited to utilize prior knowledge to improve generalization performance. In this paper, we design a novel parametric ν-support vector machine with universum data (UPar-ν-SVM). Unlabeled samples can be integrated into supervised learning by means of UPar-ν-SVM. We propose a fast method to solve the suggested problem of UPar-ν-SVM. The primal problem of UPar-ν-SVM, which is a nonconvex optimization problem, is transformed into an unconstrained optimization problem so that the objective function can be treated as a difference of two convex functions (DC). To solve this unconstrained problem, a boosted difference of convex functions algorithm (BDCA) based on a generalized Newton method is suggested (named DC-UPar-ν-SVM). We examined our approach on UCI benchmark data sets, NDC data sets, a handwritten digit recognition data set, and a landmine detection data set. The experimental results confirmed the effectiveness and superiority of the proposed method for solving classification problems in comparison with other methods. (C) 2021, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

50201 - Economic Theory

Projekt

<a href="/cs/project/GA18-04735S" target="_blank" >GA18-04735S: Nové přístupy pro relaxační a aproximační techniky v deterministické globální optimalizaci</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2021

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Applied Intelligence

ISSN

0924-669X

e-ISSN

—

Svazek periodika

52

Číslo periodika v rámci svazku

3

Stát vydavatele periodika

NL - Nizozemsko

Počet stran výsledku

21

Strana od-do

2634-2654

Kód UT WoS článku

000662820100004

EID výsledku v databázi Scopus

2-s2.0-85108106202