AN IMPROVED MULTI-VERSE OPTIMIZER FOR TEXT DOCUMENTS CLUSTERING
DOI:
https://doi.org/10.30572/2018/KJE/130203Keywords:
Multi-verse optimizer, Optimization, Test clustering, Data clustering, Neighborhood selection strategAbstract
Text document clustering (TDC) represents a key task in text mining and unsupervised machine learning, which partitions a specific documents’ collection into varied K-groups according to certain similarity/dissimilarity criterion. The multi-verse optimizer algorithm (MVO) is a stochastic population-based algorithm, which was recently introduced and successfully utilized to tackle many optimization problems that are complex. The original MVO performance is limited to the utilization of only the best solution in the exploitation phase (local search capability), which makes it suffer from entrapment in local optima and low convergence rate. This paper aims to propose a novel method of modifying the MVO algorithm called link-based multi-verse optimizer algorithm (LBMVO) to enhance the exploitation phase in the original MVO. Generally, LBMVO has outperformed or at least showed that it is profoundly competitive compared with the original MVO algorithm.
Downloads
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2022 Ammar Abasi, Ahamad Tajudin Khader , Mohammed Azmi Al-Betar
This work is licensed under a Creative Commons Attribution 4.0 International License.