Development of an Intelligent Machine Learning System for Diabetes Prediction

Abstract

In the new era of technological advancement, disease diagnosis, self-management, finding cure and predicting the possibility of one's being susceptible to some disease has become way easier. Diabetes is a silent and chronic disease mainly caused by lifestyle followed by people and all over the world millions of people are falling victim to this disease. A limited number of researches were carried out that explored the causes behind diabetes and that predicting the probability of being affected. Therefore, the objective of this research is to propose an algorithm to improve the prediction of being diabetic or non-diabetic. To attain this objective firstly, an algorithm was proposed based on Naïve Bayes with prior clustering. Second, the performance of the proposed algorithm was evaluated using 532 data related to diabetic patients. Finally, the performance of the existing Naïve Bayes algorithm was compared with the proposed. The results of the comparative study showed that the improvement in the accuracy has been made apparent for the proposed algorithm.