Bagging for Improving Accuracy of Diabetes Classification

Abstract

The quality of human life is improved by detecting diseases effectively based on the rapid development of digital image processing, internet of things and effective deep learning processing. We propose a novel application of Bootstrap Aggregation, i.e., Bagging for improving the accuracy of diabetes classification in this paper. The model of bagged logistic regression is designed to classify diabetes effectively. The ROC is used to visualize performance of the algorithm based on False_Positive_Rate (FPR) and True_Positive_Rate (TPR). The parameters performances of the proposed method are compared with the existing techniques. It is concluded that the performance of the current method with bagging is better compared to traditional techniques that do not apply any additional measures.