Dermatitis Diagnosis – Modeling and Analysis using Machine Learning

Recently, millions of people of all ages globally and animals have been affected by skin diseases, and the factors varying in the skin disorder are its severity and the identified symptoms. This painless or painful disease affecting some people, such as skin cancer, eczema, psoriasis, acne, and seborrhea dermatitis, can be temporary or permanent. These skin diseases result in early embryonic death, infertility, negative mental functioning, and low milk production. It's difficult to detect these kinds of diseases since the features affecting the skin are hair, tone, pattern, and texture. Some soft computing techniques are developed in the literature to detect these kinds of diseases, but there is a lack of accuracy in the proposed methods. This research applies the machine learning (ML) strategy, support vector machine (SVM), and image processing operations that work on the skin images dataset to detect all skin diseases in the early stages. The proposed model's accuracy, sensitivity, and specificity performance measures are 97.2%, 95.5%, and 96.1%, respectively, and obtained better performance measures than the existing methods.