Artificial rabbits optimization algorithm with automatically DBSCAN clustering algorithm to similarity agent update for features selection problems

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989100%3A27240%2F25%3A10257763" target="_blank" >RIV/61989100:27240/25:10257763 - isvavai.cz</a>

Výsledek na webu

<a href="https://link.springer.com/article/10.1007/s11227-024-06606-8#author-information" target="_blank" >https://link.springer.com/article/10.1007/s11227-024-06606-8#author-information</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/s11227-024-06606-8" target="_blank" >10.1007/s11227-024-06606-8</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Artificial rabbits optimization algorithm with automatically DBSCAN clustering algorithm to similarity agent update for features selection problems

Popis výsledku v původním jazyce

Feature selection is one of the important steps in data mining to reduce the dimensions of datasets. Due to the fact that feature selection is inherently a NP-hard problem, no deterministic algorithm has been identified to solve this problem in acceptable time. Meta-heuristic algorithms are reliable alternatives to solve such problems in acceptable time. In the literature, a large number of algorithms have been proposed to solve the feature selection problem using meta-heuristic optimization algorithms. In this work, a new feature selection algorithm based on ARO meta-heuristic algorithm and DBSCAN clustering algorithm with automatic adjustment of input parameters (ARO-DBSCAN) is proposed. Using side algorithms to improve the performance of meta-heuristic algorithms can potentially cause getting stuck in local optima. The method proposed in the work has improved the performance of the ARO meta-heuristic for the feature selection problem without increasing the probability stuck in local optima. The use of DBSCAN clustering algorithm, which is based on density, increases the exploitation of ARO in the search space while maintaining its exploration. As a result, the performance of the ARO algorithm increases significantly in feature selection problems. The proposed algorithm is compared with 8 state-of-the-art feature selection algorithms on the UCI benchmark datasets and three real-world high-dimensional datasets. Result of experiments show the better performance of ARO-DBSCAN algorithm in the appropriate execution time. Also, in high-dimensional data, the proposed method is able to significantly reduce the number of dataset features. which makes the analysis of these datasets more efficient. The source code for the algorithm being proposed is accessible to the public on https://github.com/alihamdipour/ARO-DBSCAN.

Název v anglickém jazyce

Artificial rabbits optimization algorithm with automatically DBSCAN clustering algorithm to similarity agent update for features selection problems

Popis výsledku anglicky

Feature selection is one of the important steps in data mining to reduce the dimensions of datasets. Due to the fact that feature selection is inherently a NP-hard problem, no deterministic algorithm has been identified to solve this problem in acceptable time. Meta-heuristic algorithms are reliable alternatives to solve such problems in acceptable time. In the literature, a large number of algorithms have been proposed to solve the feature selection problem using meta-heuristic optimization algorithms. In this work, a new feature selection algorithm based on ARO meta-heuristic algorithm and DBSCAN clustering algorithm with automatic adjustment of input parameters (ARO-DBSCAN) is proposed. Using side algorithms to improve the performance of meta-heuristic algorithms can potentially cause getting stuck in local optima. The method proposed in the work has improved the performance of the ARO meta-heuristic for the feature selection problem without increasing the probability stuck in local optima. The use of DBSCAN clustering algorithm, which is based on density, increases the exploitation of ARO in the search space while maintaining its exploration. As a result, the performance of the ARO algorithm increases significantly in feature selection problems. The proposed algorithm is compared with 8 state-of-the-art feature selection algorithms on the UCI benchmark datasets and three real-world high-dimensional datasets. Result of experiments show the better performance of ARO-DBSCAN algorithm in the appropriate execution time. Also, in high-dimensional data, the proposed method is able to significantly reduce the number of dataset features. which makes the analysis of these datasets more efficient. The source code for the algorithm being proposed is accessible to the public on https://github.com/alihamdipour/ARO-DBSCAN.

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

O - Projekt operacniho programu

Rok uplatnění

2025

Kód důvěrnosti údajů

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

Název periodika

Journal of Supercomputing

ISSN

0920-8542

e-ISSN

1573-0484

Svazek periodika

81

Číslo periodika v rámci svazku

1

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

34

Strana od-do

nestránkováno

Kód UT WoS článku

001351503500001

EID výsledku v databázi Scopus

—