An Aggregate Model for Prognosticate Diabetic Disease using Dissimilar Feature Selections with Upright Classification Techniques

Abstract

Our aims are to find the accuracy of classification with the normalisation in different types and the features in the techniques of selection on Diabetic Mellitus and the Pima Indian Diabetic dataset. Data Mining is the process of extraction. It extracts the previous unknown, valid and important information from the large amount of the data bases and can make the crucial decisions using the information. The classification methods are K-Nearest Neighbour and J48 decision tree can be applied to the data set of original and as well as the dataset with the pre-processed dataset. All the process of pre-processing can be applied to Pima Indian Diabetic Dataset to analyse the classification performance in terms of accuracy rate. The performance metrics is used to identify the accuracy classification is Recall, F-measure, Sensitivity and specificity, Precision, and Accuracy. The simulation is done by R tool.