A Machine Learning Approach to Predict Skin Diseases and Treatment Recommendation System

In medical science, dermatology is the department dedicated to identify, diagnose, and suggest a treatment for skin disease. Diseases that occur on human skin is mostly unpredictable and tough to analyze, in recent days, the occurrence of skin diseases increased unprecedently. Skin diseases are usually caused by bacteria or other related infections. Some skin diseases like ringworm, yeast infection, allergies etc. will constantly increase and spread over the skin. This type of disease should be identified in its earlier stage to avoid spreading. The identification can be done by considering some factors like clinical parameters, which are considered for identifying the disease. The possible skin diseases are dermatitis for the age group 0-5 years, warts affect 6-11 years old children, and acne vulgaris affect 12-16 years. Dermatomyositis is a type of skin disease that affects children aged 5-15 and adults aged 40-60. Due to unawareness, patients usually ignore the early symptoms. This necessitates the need for skin disease prediction. This research work predicts the type of disease by using the clinical parameter of a patient, which is achieved by using various machine learning classification algorithms. Maximum accuracy of 98.4% is achieved by the Naive Bayes algorithm. The Graphical User Interface (GUI) was developed for easy utilization of this prediction module for both doctors and dermatologists.