Novel Machine Learning Ensemble Models for Active Diabetes Diagnosis

Abstract

Now that diabetes is on the rise as a silent killer, it will have catastrophic consequences if preventative measures are not implemented soon enough. Given that diabetes cannot be cured after it has been diagnosed in a patient, early identification of the illness is critical in diabetes. Doctors are finding it more difficult to do manual detection as the number of patients grows on a daily basis. We can do some automobile detection using tools like Machine Learning. This issue of diabetes diagnosis has been the subject of much study up to this point. Using two ensemble machine learning algorithms like Random Forest and GBDT, this study does predictive analysis. Pima Indians Diabetes Dataset, which includes data of diabetes patients, was used in this study to conduct different experiments. The findings are presented in this article. This article also addresses the significance of output interpretability in the healthcare sector and explains how providing output interpretability together with the machine learning model’s output on the patient record would assist physicians in real-time (in the future).