A Binary Multi-Objective Chimp Optimizer With Dual Archive for Feature Selection in the Healthcare Domain

Abstract

Medical datasets frequently include vast feature sets with numerous features that are related to one another. As a result, the curse of dimensionality affects learning from a medical dataset to discover significant characteristics, making it necessary to minimize the feature set. Feature selection (FS) is a major step in classification and also in reducing the dimension. This study attempts a novel Binary Multi-objective Chimp Optimization Algorithm (BMOChOA) with dual archive and k-nearest neighbors (KNN) classifier for mining relevant aspects from medical data. In this research, 12 versions of BMOChOA are implemented based on the group information and types of chaotic functions used. The best Pareto front obtained from suggested BMOChOA variations is compared with three benchmark multi-objective FS methods by taking 14 popular medical datasets of variable dimensions. By analyzing the experimental outputs using four multi-objective performance evaluators, it is found that the proposed FS method is superior in finding the best trade-off between the two objective functions: the number of features and classification performance.