EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Multi-objective enhanced binary bat algorithm for unsupervised feature selection with classification validation

Alhassan Adamu, Isah Tajiri Sagagi, Salisu Mamman Abdulrahman, Umma Abubakar, Bashir Salisu Abubakar, Ali Olow Jimale

Journal of Statistical Sciences and Computational Intelligence · 29.06.2026 · 2 (2) · 691-700

DOI: 10.64497/jssci.269 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Statistical Sciences and Computational Intelligence, online erschienen am 29.06.2026. Verfasst von Alhassan Adamu und 5 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Feature selection is a vital preprocessing step for reducing dimensionality and improving model generalization, especially in high-dimensional datasets. This paper presents a Multi-Objective Enhanced Binary Bat Algorithm (MO-EBBA) for unsupervised feature selection, integrating a novel crossover operator and adaptive pulse rate mechanism. Unlike traditional single-objective approaches, MO-EBBA simultaneously optimizes two conflicting objectives: maximizing clustering quality (measured via Silhouette Coefficient) and minimizing the number of selected features. The selected features are further validated through multiple classifiers (SVM, k-NN, Decision Tree, and Random Forest), using supervised metrics such as accuracy, F1-score, precision, and recall. We conduct experiments on several benchmark datasets, demonstrating that MO-EBBA achieves superior trade-offs between feature reduction and classification performance compared to existing state-of-the-art methods. Also, we conducted statistical tests to confirm the significance of the improvements of the new method.

Zitieren

APA

Adamu, A., Sagagi, I. T., Abdulrahman, S. M., Abubakar, U., Abubakar, B. S., & Jimale, A. O. (2026). Multi-objective enhanced binary bat algorithm for unsupervised feature selection with classification validation. Journal of Statistical Sciences and Computational Intelligence, 2(2), 691-700. https://doi.org/10.64497/jssci.269

Vancouver

Adamu A, Sagagi IT, Abdulrahman SM, Abubakar U, Abubakar BS, Jimale AO. Multi-objective enhanced binary bat algorithm for unsupervised feature selection with classification validation. Journal of Statistical Sciences and Computational Intelligence. 2026;2(2):691-700. doi:10.64497/jssci.269

BibTeX
@article{Adamu2026,
  author = {Adamu, Alhassan and Sagagi, Isah Tajiri and Abdulrahman, Salisu Mamman and Abubakar, Umma and Abubakar, Bashir Salisu and Jimale, Ali Olow},
  title = {Multi-objective enhanced binary bat algorithm for unsupervised feature selection with classification validation},
  journal = {Journal of Statistical Sciences and Computational Intelligence},
  year = {2026},
  volume = {2},
  number = {2},
  pages = {691--700},
  doi = {10.64497/jssci.269}
}
Kostenlos starten →