Analysis of machine learning methods to determine the best data analysis method for diabetes prediction

Abstract

Chronic diabetes results from the body's inability to produce adequate insulin. It is an incurable but treatable disease. The experiment conducted in this study aims to analyze different machine learning methods like Stochastic Gradient Descent, Support Vector Machine, Logistic Regression, and CN2 Rule using the Orange data mining software and use them for diabetes prediction based on the PIMA Indian Diabetes Dataset. Utilizing different performance criteria like Accuracy, Precision, Recall, and F1-Score, these approaches were examined and evaluated. The best outcome was obtained by CN2 Rule Induction, achieving an accuracy score of 80.7% which shows that this method is the most suitable for diabetes prediction compared to the other three models.