International Journal of Computer Science & Information Technology (IJCSIT) Vol 7, No 3, June 2015 DOI:10.5121/ijcsit.2015.7310 115 MULTI-PARAMETER BASED PERFORMANCE EVALUATION OF CLASSIFICATION ALGORITHMS Saurabh Kr. Srivastava 1 and Sandeep Kr. Singh 2 1 Research Scholar, Department of Computer Sc. & Engineering, JIIT University, Noida, India 2 Assistant Professor, Department of Computer Sc. & Engineering, JIIT University, Noida, India ABSTRACT Diabetes disease is amongst the most commondisease in India. It affects patient’s health and also leads to other chronic diseases. Prediction of diabetes plays a significant role in saving of life and cost.Predicting diabetes in human body is a challenging task because it depends on several factors. Few studies have reported the performance of classification algorithms in terms of accuracy. Results in these studies are difficult and complex to understand by medical practitioner andalso lack in terms of visual aids asthey are presented in pure text format. This reported survey uses ROC and PRC graphical measures to improve understanding of results. A detailed parameter wise discussion of comparison isalso presented which lacks in other reported surveys. Execution time, Accuracy, TP Rate, FP Rate, Precision, Recall, F Measure parameters are used for comparative analysis and Confusion Matrix is prepared for quick review of each algorithm. Ten fold cross validation method is used for estimation of prediction model. Different sets of classification algorithms are analyzed on diabetes datasetacquired from UCI repository. KEYWORDS Data mining, Diabetes, UCI repository, Machine Learning. 1. INTRODUCTION AND LITERATURE REVIEW Diabetes invites other chronic diseases. It is primarily associated with an increase of blood glucose. Two common categories of diabetes are Type-1 and Type-2. Type-1 diabetes occurs when pancreas fails to produce sufficient insulin while Type-2 occurs when body cannot effectively consume the insulin produced. Diabetes confirmation requires a lot of examination that affects time and cost both. Prior studies show that classification algorithms have been effectivelyused in prediction of diabetes. Several studies and results are reported using data mining techniques in healthcare for classification in medical databases. In the context of that J.W.Smithetal.[1] have introduced adaptivelearningroutinethatexecutes digital analogyofperceptionscalled ADAP. The algorithm uses 576 training instances and classification accuracyis76%onthe remaining192instances. K.Srinivaset al.[4] studied theapplicationsofDataminingTechniquesin healthcareand have reported predictionof heart attacks using clinical database.Elmakolce et al.[5] has given the glimps ofdifferentdataminingtechniquesutilization. AshaRajkumaret al.[2]and A.Khemphlilaet al.[3] discussedvariousdataminingtechniques fordiagnosisof certainlife threateningdiseases. HuyNguyen A.P. et al.[6] proposeda newHomogeneity- BasedAlgorithmthatdeterminesoverfittingandovergeneralizationbehaviorofclassification.Recently K.R. Lakshmiet al.[10] and Karthikeyini.V. et.al [8,9] discusseddatamining algorithms performancebasedupon theircomputingtime and precisionvalue.