Enhanced Multi-Task Learning Using Optimization Methods

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F44555601%3A13440%2F25%3A43899196" target="_blank" >RIV/44555601:13440/25:43899196 - isvavai.cz</a>

Výsledek na webu

<a href="https://link.springer.com/article/10.1007/s41060-025-00859-z" target="_blank" >https://link.springer.com/article/10.1007/s41060-025-00859-z</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/s41060-025-00859-z" target="_blank" >10.1007/s41060-025-00859-z</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Enhanced Multi-Task Learning Using Optimization Methods

Popis výsledku v původním jazyce

This paper presents an improved multi-task twin support vector machine method utilizing Universum data, referred to as I-UMTSVM. A significant advancement of I-UMTSVM over UMTSVM is the addition of a regularization term, which facilitates the application of the structural risk minimization principle. This modification captures the essence of statistical learning theory and leads to enhanced classification performance. Two approaches are proposed for solving the I-UMTSVM problem. The first is a dual formulation of I-UMTSVM, which involves solving a quadratic programming problem. The second, NI-UMTSVM, is a Newton-based approach that addresses I-UMTSVM in the primal space. In NI-UMTSVM, the constrained optimization problems of I-UMTSVM are transformed into unconstrained ones, and a generalized Newton&apos;s method is introduced to solve them effectively. The efficiency of the proposed methods is demonstrated through numerical experiments on various benchmark multi-task data sets. These experiments provide evidence of the effectiveness and strong performance of the proposed approaches in multi-task learning tasks.

Název v anglickém jazyce

Enhanced Multi-Task Learning Using Optimization Methods

Popis výsledku anglicky

This paper presents an improved multi-task twin support vector machine method utilizing Universum data, referred to as I-UMTSVM. A significant advancement of I-UMTSVM over UMTSVM is the addition of a regularization term, which facilitates the application of the structural risk minimization principle. This modification captures the essence of statistical learning theory and leads to enhanced classification performance. Two approaches are proposed for solving the I-UMTSVM problem. The first is a dual formulation of I-UMTSVM, which involves solving a quadratic programming problem. The second, NI-UMTSVM, is a Newton-based approach that addresses I-UMTSVM in the primal space. In NI-UMTSVM, the constrained optimization problems of I-UMTSVM are transformed into unconstrained ones, and a generalized Newton&apos;s method is introduced to solve them effectively. The efficiency of the proposed methods is demonstrated through numerical experiments on various benchmark multi-task data sets. These experiments provide evidence of the effectiveness and strong performance of the proposed approaches in multi-task learning tasks.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2025

Kód důvěrnosti údajů

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

Název periodika

International Journal of Data Science and Analytics

ISSN

2364-415X

e-ISSN

2364-4168

Svazek periodika

2025

Číslo periodika v rámci svazku

"necislovano"

Stát vydavatele periodika

GB - Spojené království Velké Británie a Severního Irska

Počet stran výsledku

22

Strana od-do

"nestrankovano"

Kód UT WoS článku

001532968200001

EID výsledku v databázi Scopus

—