Skin Disease Predication And Analyzing Using Naive Bayes Classification Algorithms

Abstract

In modern times medicine, skin disorders are widely recognized as prevalent illnesses in the human population. Skin cancer is a life-threatening cancer that poses a significant public health problem, with early detection being crucial for effective treatment and survival. The Severity of skin cancer and the rapid count of affected people make it necessary to introduce an automatic detection scheme. Generally, analysing and identifying skin disease in a short time is the most complex and challenging task. Several machine learning (ML) methods are introduced to achieve this. However, still fulfilling the skin cancer diagnosis is not accomplished completely. To achieve this, we proposed a machine learning model using naive bayes  with ROC to predict skin disease with maximum accuracy. The proposed naive bayes  is based on similar features and classifies several stages. The performance obtained by the naive bayes  is compared with  ensemble classifiers and CNN with several evaluation metrics. The analysis shows that the accuracy attained by the proposed naive bayes  is 98.5 % far better than the others in terms of classification and accuracy..