Monkeypox diagnosis based on Dynamic Recursive Gray wolf (DRGW) optimization

Abstract

The increasing spread of monkeypox recently requires more study, investigation, and analysis. Not only to understand the different factors that contribute to the spread and transmission of the disease, but also to devise rules by which the disease can be diagnosed more accurately as the clinical presentation of monkeypox is very similar to that of smallpox. This paper presents a simple but effective strategy for diagnosing monkeypox, which is called Accurate Monkeypox Diagnosing Strategy (AMDS). The proposed AMDS consists of two successive phases, which are; (i) Feature Selection Phase (FSP) and (ii) Classification Phase (CP). During the FSP, the most beneficial features for diagnosing monkeypox are selected; afterwards the CP is applied during which the actual diagnosis takes place. The main contribution of AMDS relies on a proposed improvement of the Traditional Gray Wolf Optimization (TGWO), which is identified as Dynamic Recursive Gray Wolf Optimization (DRGW) and is applied in the FSP to select the most influential features. DRGW solves two of the most effective drawbacks of TGWO, which are; (i) inability to share good positions among wolves’ pack and (ii) the lack of an accurate mechanism to locate the potential prey. AMDS has been experimentally evaluated compared to other competitors of recent feature selection techniques, in which CP has been employed using different well-known classifiers such as; Support Vector Machines (SVM), Naïve Bayes (NB), K-Nearest Neighbors (KNN), and Deep Neural Network (DNN). Experimental results conclude that AMDS promotes the diagnosing performance in terms of recall, accuracy, error, precision, and F score. This ensures that the proposed AMDS can operate efficiently in the disease diagnosis, which in turn ensures the effectiveness of the proposed DRGW.