Prediction and Comparison of Diabetes with Logistic Regression, Naïve Bayes, Random Forest, and Support Vector Machine

Abstract

Diabetes, is a major metabolic disorder, caused by abnormally elevated blood glucose or sugar concentration. Increased glucose levels can cause high harm to the heart, kidney, eyes, and blood vessels. According to the WHO, over 422 million people suffer it. Every year, this disease causes more than 1.5 million deaths. It is rampant in low and middle-income countries. Modern machine learning (ML) techniques improve predictions and performance. This study focuses on ML classification algorithms in a diabetes dataset for reliably predicting diabetes using Python. Six ML methods are used, i.e., random forest, logistic regression, XG boost, support vector machines, Naive Bayes, and KNN. In contrast to other ML algorithms, random forest had the highest accuracy of 97.02%.