Ensemble Technique for Diabetic Precision Medicine Classification

Abstract

It is precisely 100 years after giving an insulin shot to a human, which created a revolution in diabetes treatment. Since the fate of diabetes patients has changed in humankind. In this regard, the exact dosage and same set of medicine for diabetes patients may not fit. Thus, there is a requirement for Precision medicine which can change the treatment of diabetes. There is a requirement for building automated intelligent systems to recommend precision medicine that can help practitioners. This paper discusses precision medicine, which has a complete schema for deriving Precision medicine from Big data. In our proposed schema, the component Intelligent Precision Medicine Engine, a new phase is added to filter the non-diabetic patient records to be processed by the Recommender engine, which would reduce the computational energies. With this object, a multi-layered bagging technique is used to classify with the best result nearing 96% with the UCI machine learning dataset for diabetes. Three layers of classification multi-models with majority voting are designed, and results are discussed.