Evaluation of Ensemble Prediction Techniques on Electronic Health Data

Electronic health care data plays fundamental role in decision support system of ensemble learning techniques for disease determination. This paper exhibits a new ensemble prediction model of different ensemble classifiers are fused together by base classifier. The classification accuracy of the proposed prediction model is improved by efficient pre-processing and ten-fold validation method. The prediction model is modelled as an ensemble fusion and misclassification of instances are reported. The prediction of the proposed ensemble fusion model has achieved better prediction when compared to other existing model. The experiment for evaluating the proposed ensemble fusion was carried out using the standard datasets viz UCI repository Pima Indian diabetes and breast cancer datasets. The misclassification instances of the proposed model was decreased.