Sparse least squares K-SVCR multi-class classification

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F44555601%3A13440%2F24%3A43898581" target="_blank" >RIV/44555601:13440/24:43898581 - isvavai.cz</a>

Result on the web

<a href="https://jnva.biemdas.com/archives/2507" target="_blank" >https://jnva.biemdas.com/archives/2507</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.23952/jnva.8.2024.6.07" target="_blank" >10.23952/jnva.8.2024.6.07</a>

Result language

angličtina

Original language name

Sparse least squares K-SVCR multi-class classification

Original language description

This paper introduces a novel model, the sparse least squares K-class support vector classificationregression with adaptive p-norm (PLSTKSVC), to tackle challenges in multi-class classification. Leveraging a &quot;1-versus-1-versus-rest&quot; structure, PLSTKSVC dynamically adjusts the parameter p based on the data, enabling an adaptive learning framework. By incorporating cardinality-constrained optimization, the model seamlessly integrates feature selection and classification. Although the p-norm is non-convex for 0 &lt; p&lt;1, PLSTKSVC efficiently addresses the associated optimization via linear systems of equations. PLSTKSVC offers several advantages, including simultaneous feature selection and classification, robust theoretical foundations, algorithmic efficiency, and strong empirical validation. The model?s theoretical contributions include lower bounds on non-zero solution entries and upper bounds on the optimal solution norm. Experimental results on multi-class classification datasets highlight the superior performance of PLSTKSVC, establishing it as a significant advancement in machine learning.

Czech name

—

Czech description

—

Type

J_imp - Article in a specialist periodical, which is included in the Web of Science database

CEP classification

—

OECD FORD branch

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

Project

—

Continuities

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Publication year

2024

Confidentiality

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

Name of the periodical

Journal of Nonlinear and Variational Analysis

ISSN

2560-6921

e-ISSN

2560-6778

Volume of the periodical

8

Issue of the periodical within the volume

6

Country of publishing house

CA - CANADA

Number of pages

19

Pages from-to

953-971

UT code for WoS article

001374806800007

EID of the result in the Scopus database

2-s2.0-85208017159