Archives

Developing A Conventional Ensemble Model with Weighted Vote for Diabetes Prediction


S. Sathurthi and Dr.K. Saruladha
Abstract

Ensemble learning is one of the necessary supervised learning techniques for developing disease prediction model for diagnosis of diabetes. Diabetes Mellitus is one of the important health issues in the world. This paper exhibits a new ensemble prediction model of various ensemble and base classifiers are fused together by weighted voting method. The efficiency of the system has been evaluated by tenfold validation method and the results have been calculated using percentage of correctly classified and incorrectly classified instances. In ensemble based prediction model for diabetes have been compared with base classifier fusion model. Finally, the result indicate that conventional ensemble prediction model can achieve better predictive performance. This system has been experimented in Python language with UCI Machine Learning Repository diabetes dataset.

Volume 11 | 04-Special Issue

Pages: 1130-1137