Binary Banyan Tree Growth Optimization: A Practical Approach to High-dimensional Feature Selection

<p dir="ltr">High-dimensional feature spaces in Scientific and Technical Service Resources (STSR)<br>classification present significant challenges, including increased computational costs and<br>diminished accuracy. Identifying an optimal subset of features from raw text vectors is thus critical<br>for effective data classification. This paper introduces a novel metaheuristic algorithm called Binary<br>Banyan Tree Growth Optimization (BBTGO), specifically designed for high-dimensional feature<br>selection (FS). Inspired by the unique growth patterns of the banyan tree, BBTGO leverages a<br>combination of innovative Boolean vectors, including rooting, multi-trunk, and adjustment operator,<br>along with a perturbation phase to enhance the search efficiency and reduce feature dimensionality.<br>These operators enhance the search for promising regions and reduce features by utilizing the<br>optimal solutions clustered within subgroups. Furthermore, BBTGO incorporates a dynamic<br>adjustment mechanism that periodically activates different growth operators to meet the search<br>demands of high-dimensional space. We rigorously evaluate the exploration and exploitation<br>capabilities of BBTGO through comprehensive statistical analyses of various performance metrics.<br>The proposed method demonstrates superior results on 12 high-dimensional benchmark datasets<br>and is successfully applied to feature selection in STSR text classification tasks. Experimental<br>results show that BBTGO significantly outperforms existing methods in terms of classification<br>accuracy, selected features, convergence speed, and processing time. These results underscore the<br>potential of BBTGO as a robust and versatile solution for high-dimensional FS, with broad<br>applicability to real-world classification challenges.</p>