Predicting the Effect of Diabetes on Kidney using Classification in Tanagra?
Divya Jain, Sumanlata Gautam
Journal Title:International Journal of Computer Science and Mobile Computing - IJCSMC
As numerous data mining tools & techniques continue to develop along with the healthcare domain, the applications of data mining in healthcare sector will undoubtedly play a growing role in the world. There exist a large number of useful techniques in data mining like classification, association, clustering, regression etc. that helps to discover new trends in huge healthcare databases. This paper comes out with the application of classification and prediction technique to find the effect of diabetes on kidney. The implementation is done using the application of C4.5 Algorithm in Tanagra. This paper demonstrated the utility of classification on a dataset containing records of both diabetic and non-diabetic patients using data mining tool Tanagra. Using Kidney Function Tests (KFT), the effect of diabetes on kidney is determined. After comparing manually the result of Tanagra with the actual output, we discovered that Tanagra is very near to the output. Finally, the performance of classifier in Tanagra is evaluated in terms of recall, precision and error rate.