Abstract:Classification of medical data is largely dependent on the effective identification of key features of the data that can be used to aid in the diagnosis of related diseases. This goal can be achieved through feature selection methods that endeavor to get rid of redundant and irrelevant features to ameliorate classification accuracy. This is the aim of this work where a new meta-heuristic, referred to as snake optimizer, was adopted for the purpose of boosting the performance of existing feature selection methods. This optimizer may smoothly fall into local optimal solutions, which may present weak search performance and slow convergence speeds in handling feature selection problems. On this basis, this paper presents three improved adaptive versions of this optimizer, each of which has increased search performance over the basic optimizer. This optimizer was improved using three mathematical models named exponential, power, and delayed S-shaped, to create three methods, referred to as exponential, power, and delayed S-shaped snake optimizers, respectively. These proposed versions were also matured to have more balance between exploration and exploitation aspects. Then, binary variants of these optimizers were evolved to solve feature selection problems using the k-nearest neighbor classifier. To verify the efficacy of these binary optimizers, 24 datasets were used, and then compared with other feature selection optimizers. The experimental results obviously manifested the efficiency of the proposed optimizers in realizing the optimal feature set by achieving utmost accuracy and minimal number of features in the majority of the studied datasets. The proposed binary power snake optimizer outperformed all other competitors in 13, 10, 8, 8, and 12 datasets in respect of classification accuracy, number of chosen attributes, specificity, sensitivity, and fitness scores, respectively. Out of the 24 datasets taken into consideration, the results on 12, 6, and 8 datasets, respectively, showed that this proposed optimizer presented performance scores of more than 90% in respect of sensitivity, accuracy, and specificity metrics.

An improved Differential evolution with Sailfish optimizer (DESFO) for handling feature selection problem

A Novel Feature Selection Method Based on Extreme Learning Machine and Fractional-Order Darwinian PSO.

Differential Evolution Based Feature Selection: A Niching-based Multi-objective Approach

A Hybrid Binary Dragonfly Algorithm with an Adaptive Directed Differential Operator for Feature Selection

A differential evolution framework based on the fluid model for feature selection

Differential Evolution With Duplication Analysis for Feature Selection in Classification

An Enhanced Snake Optimizer Algorithm for Feature Selection

Opposition-based Moth-Flame Optimization Improved by Differential Evolution for Feature Selection

New Feature Selection Paradigm Based On Hyper-Heuristic Technique

An improved self-adaptive differential evolution algorithm and its application

Reinforcement Learning-Based Multi-Objective Differential Evolution Algorithm for Feature Selection

An Enhanced Dynamic Differential Annealed Algorithm for Global Optimization and Feature Selection

The monarch butterfly optimization algorithm for solving feature selection problems

An adaptive hybrid differential evolution algorithm for continuous optimization and classification problems

MFeature: Towards high performance evolutionary tools for feature selection

Feature Selection Using Diversity-Based Multi-objective Binary Differential Evolution

SFE: A Simple, Fast and Efficient Feature Selection Algorithm for High-Dimensional Data

Improved versions of snake optimizer for feature selection in medical diagnosis: a real case COVID-19

Feature Selection Based on Improved White Shark Optimizer

Multiobjective Differential Evolution for Feature Selection in Classification

An Efficient Parallel Reptile Search Algorithm and Snake Optimizer Approach for Feature Selection