An Efficient Hybrid Classification Approach for COVID-19 Based on Harris Hawks Optimization and Salp Swarm Optimization

Abstract

Feature selection can be defined as one of the pre-processing steps that decreases the dimensionality of a dataset by identifying the most significant attributes while also boosting the accuracy of classification. For solving feature selection problems, this study presents a hybrid binary version of Harris Hawks Optimization algorithm (HHO) and Salp Swarm Optimization (SSA) (HHOSSA). The HHOSSA was tested against two well-known optimization algorithms, the Whale Optimization Algorithm (WOA) and the grey wolf optimizer (GWO), utilizing 280 2D X-ray images from the Posteroanterior (PA) chest view dataset for normal and covid-19 patients. A total of three performance metrics (Recall, Precision, F1) were employed in the studies with Support vector machines (SVMs), k-Nearest Neighbor (KNN), and XGBoost as classifiers. The suggested algorithm outperforms SVM by 96%, and two classifiers, XGboost and KNN, by 98%.