Machine Learning in Healthcare Diagnosis System using Density Based Clustering with Logistic Regression Classification Model

Abstract

Diabetes is considered as a commonly occurring disease affecting people under all ages over the globe. Due to massive increase in number of patients being affected by diabetes, it is essential to design an automated disease diagnosis model for diabetes. This paper introduces a new machine learning based clustering with classification model to diagnose the presence of diabetes. The proposed model involves a density based clustering (DBC) technique with LR (LR) classifier called DBC-LR model for diabetes diagnosis. The presented DBC-LR model comprises a set of four stages namely preprocessing, feature reduction, clustering and classification. The validation of the DBC-LR model takes place using Pima Indian Diabetes dataset and the results are validated under several aspects. The clustered output of the DBC-LR model on the applied dataset and the classification outcome ensures the betterment of the proposed model over the compared methods. The DBC-LR model has resulted to superior outcome with the maximum accuracy of 98.3%.