Classification of Diabetes Dataset using KNN Classifier and Attribute Selection through Bees Algorithm

Abstract

Diabetes is considered as a one of the chronic disease it cause an increase of sugar in the human blood. The prime objective of this research work is provides better classification accuracy through reduced set of attributes. In this work, first the classification is carried out with entire dataset, and then second the classification is carried out through reduced dataset i.e., with reduced number of attributes. The proposed system has evaluated through three performance metrics such as Detection Rate (DR), Accuracy Rate (AR) and False Positive Rate (FPR). The proposed system provides good AR, DR and FPR what we have obtained through entire dataset.