A Review of Ensemble Machine Learning Approach in Prediction of Diabetes Diseases

Abstract

Data mining techniques improve efficiency and reliability in diabetes classification. Machine learning techniques are applied to predict medical dataset to safe human life. The large set of medical dataset is accessible in data warehousing which used in the real time application. Currently Diabetes Diseases (DD) is among the leading cause of death in the world. Data mining techniques are used to group and predict symptoms in medical dataset by different examiners. Data set from Pima Indian Diabetes Dataset (PIMA) were utilized to compare results with the results from other examiners. In this system, the most well known algorithms; K-Nearest Neighbor (KNN), Na�ve Bayes (NBs), Random Forest (RF) and J48 are used to construct an ensemble model. The experiment�s result reveals that an ensemble hybrid model increases the accuracy by combining individual techniques in to one. As a result, the model serves to be useful by doctors and Pathologist for the realistic health management of diabetes.