Binary Magnificent Frigatebird Optimization: A Next-Generation Algorithm for Enhanced Feature Selection and Classification in High-Dimensional Data

View at publisher Summarise with VIRAT
Authors: D. Mansoor Hussain, B. Arunkumar, S. Suresh Kumar, P. Mathiyalagan, S. Narayanasamy, R Kaviyaraj
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The rapid advancement in data science and big data has led to the accumulation of large and complex datasets, presenting challenges in feature selection for improving classification accuracy. Motivated by frigatebirds’ kleptoparasite nature, the Magnificent Frigatebird Optimization (MFO) method excels in continuous optimization tasks but does not address binary optimization challenges such as feature selection. To overcome this limitation, we propose the Binary Magnificent Frigatebird Optimization (BMFO) algorithm, which adapts the MFO algorithm for binary search spaces using novel position update techniques. Our study evaluates the performance of BMFO through experiments on four datasets: Breast Cancer, Central Nervous System, OSCC, and Ovarian Cancer. The BMFO approach was tested against the S-shaped and V-shaped variants of MFO, focusing on feature selection and classification accuracy. Results demonstrate that BMFO consistently outperforms traditional MFO methods, with the V-shaped variant showing significant improvements in classification accuracy and feature selection in datasets such as Ovarian Cancer. Specifically, BMFO achieved accuracies of 0.62 in Ovarian Cancer, surpassing the S-shaped variant’s accuracy of 0.52. The BMFO approach also achieved the lowest fitness values across all datasets, highlighting its superior performance in optimizing feature selection and enhancing model accuracy. The findings suggest that BMFO, particularly with the V-shaped update mechanism, offers a robust solution for feature selection in high-dimensional data, providing a more effective and efficient approach compared to existing methods.

Fields of Study & Subjects

Metaheuristic Optimization Algorithms Research Artificial Immune Systems Applications Imbalanced Data Classification Techniques

Publication details

Journal / source
CRC Press eBooks
Pages
98–104
Publication date
31 March 2026
Language
EN
Document type
Book chapter