Liable Characteristics Measure and Anticipate the Diabetes Disease Using Machine Learning Tools

Abstract

Diabetes is a cardiovascular disease. It is not only an epidemic in Bangladesh but also in the whole world that is increasing rapidly. At an early period of human life, machine learning techniques are used to predict diabetes datasets. In our research paper, we use the Pima diabetes dataset from the Kaggle UCI machine learning data repository. For diabetic patients and doctors, machine learning techniques are both cost-effective and time-saving. We apply KNN, Nave Bayes, Random forest, Support vector machine, Simple logistic, and J48 to Pima datasets. Besides these algorithms, we may develop an ensemble (Vote) hybrid model with WEKA software by combining individual methods that provide the best performance and accuracy. Also, try to make a comparison among all machine learning tool’s accuracy and performance with the proposed ensemble model.