Machine Learning Approaches for Predicting Monkeypox Disease

The spread of monkeypox virus around the world to more than 40 countries outside Africa has developed into a critical public health issue. Early detection of monkeypox becomes challenging since doctors must rule out chickenpox and measles symptoms. The detection of monkeypox lesions through computers functions as an essential monitoring instrument for identifying possible infection cases when PCR testing is challenging to access. A combination of machine learning algorithms could perform automatic lesion identification although this approach requires enough available training data. A standard data collection for monkeypox lesion images does not exist at present. The research work delivers Monkeypox Skin Lesion Dataset (MSLD) that combines photos of monkeypox lesions alongside pictures of chickenpox and measles skin injuries. These images derive from news platforms in addition to blogs and case reports that people can access through the public domain. The data augmentation process occurs during the execution of a three-fold validation experiment. Support Vector Machines (SVM) serve as part of various pre-trained machine learning models during disease classification at the subsequent stage. The system operationalizes three separate models as an ensemble modeling approach. Staff members construct a web-based application that permits remote screening assessments for monkeypox disease through the internet interface. The good results from our initial dataset require careful consideration because expanded observational research on larger datasets will improve prediction accuracy.