Investigation of the Performance of Classification Methods in Hepatitis Decision Support Systems

In addition to developing information technologies, medical decision support systems are used to make decisions more effective, faster, more accurate and easier to make decisions in medicine and healthcare. With data mining, which is an important part of medical decision support systems, it is possible to obtain valuable knowledge among large-scale data, to show the linkage of data with each other, and to foresee the future. In this study, the success of classification models, which are predictive models of data mining used in medical decision support systems, has been investigated in determining whether hepatitis disease is fatal in terms of certain determinants. For this purpose, data set of hepatitis disease was analyzed and C4.5 decision tree, k-NN (IBK), Naive Bayes and multilayer artificial neural network (MLP) classification methods were applied to preprocessed data set through WEKA program.When the applied methods success rates of the hepatitis-diagnosed were compared, although all methods have acceptable performance, It was determined that the most successful C4.5 decision tree classification method with success rate of 89.03%.The results of the study demonstrated that the C4.5 Decision Tree algorithm could be used effectively as a medical decision support system for diagnosing hepatitis.