Dermatological Disorder Detection Using Machine Learning

Abstract

Skin diseases are considered one of the biggest scientific troubles in 21st century because of its especially complex and luxurious prognosis with problems and subjectivity of human interpretation. In cases of deadly illnesses like Melanoma prognosis in early tiers play a critical part in determining the possibility of getting cured. The software of automated strategies will assist in early diagnosis specifically with photographs with variety of analysis. Hence, in this system we present a completely automated machine of skin sickness recognition via lesion images, a device intervention in evaluation to traditional clinical personnel based detection. This system is designed into 3 levels compromising of statistics series and augmentation, designing version and subsequently prediction of disease. This proposed system uses more than one AI algorithms like Convolutional Neural Network and naive Bayes classifier and amalgamated it with image processing tools to shape a higher shape, leading to better accuracy.