A Hybrid Feature Selection Method BFSSFBS and Bayesian net for Diagnosis of Dermatology Diseases
Abstract
It is discernible that usage of machine learning methods in disease diagnosis has been increasing gradually. In this study, diagnosis of different dermatology diseases, which is a very common and important disease, was conducted with such a machine learning system. In this research detection has been made on dermatology diseases using best first search(BFS), sequential floating backward search(SFBS) and Bayesian net. The approach system has three stages. In the first stage, dimensions of erythemato squamos dataset that has 35 features is reduced to 19 features using best first search. In the second stage, sequential floating backward search was utilized as a pre-processing step before the main classifier. Then, in the third stage, bayesian net is used as a classifier. This method has been named as BFSSFBS. In this erythemato squamous dataset is taken from the UCI machine learning database. The obtained classification accuracy of proposed model was 99.31% and it was very promising with regard to the other classification applications in the literature for this problem.