An artificial bee bare-bone hunger games search for global optimization and high-dimensional feature selection.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• High-dimensional genetic data presents challenges in identifying representative genes and reducing dimensionality.
• Effective gene selection is crucial for minimizing computational costs and improving classification precision.

Purpose of the Study:

• To design a novel wrapper gene selection algorithm, artificial bee bare-bone hunger games search (ABHGS), to address challenges in high-dimensional genetic data analysis.
• To enhance the hunger games search (HGS) algorithm by integrating an artificial bee strategy and a Gaussian bare-bone structure.

Main Methods:

• The proposed ABHGS algorithm combines the hunger games search (HGS) with an artificial bee strategy and a Gaussian bare-bone structure.
• Performance evaluation involved comparing ABHGS against HGS, its embedded strategies, six classic algorithms, and ten advanced algorithms using CEC 2017 functions.

Main Results:

• The ABHGS algorithm demonstrated superior performance compared to the original HGS and other benchmark algorithms.
• Experimental results showed an increase in classification accuracy and a reduction in the number of selected features using ABHGS.

Conclusions:

• The proposed ABHGS algorithm is effective for gene selection in high-dimensional genetic data.
• ABHGS exhibits significant engineering utility in spatial search and feature selection, outperforming existing methods.