Efficient Skin Lesion based Classification System for Monkeypox Detection using VGG16 and Ensemble Learning

Monkeypox is a recently emerged disease outbreak that has impacted the health of many individuals. This study focuses on developing a skin lesion-based classification system for monkeypox disease. Many recent works suggest that the direct implementation of machine learning (ML) and deep learning (DL) approaches for this task has not resulted in high accuracy. Therefore, this paper proposes an efficient technique that employs a hybrid of VGG16-based feature engineering, class balancing, and ensemble learning for the detection of monkeypox skin lesions. Specifically, DL-based feature extraction is applied, but the classification is carried out through ensemble learning techniques. The results demonstrate that the random forest, extra tree, and histogram gradient descent methods achieved better performance in terms of different performance metrics.