Performance Analysis of Machine Learning Algorithms and Feature Selection Methods on Hepatitis Disease

Authors : Ebru AYDINDAĞ BAYRAK, Pınar KIRCI, Tolga ENSARİ

Pages : 135-138

View : 26 | Download : 9

Publication Date : 2019-12-23

Article Type : Conference Paper

Abstract :In this study, some machine learning classification techniques are applied on Hepatitis data set acquired from UCI Machine Learning Repository. Naïve Bayes Classifier, Logistic Regression and J48 Decision Tree are used as classification algorithms and they have been compared according to filter-based feature selection methods. For filter-based feature selection, Cfs Subset Eval, Info Gain Attribute Eval and Principal Components have been used and the performance of them is evaluated in terms of precision, recall, F-Measure and ROC Area. Among the all used classification algorithms, Naïve Bayes Classifier has higher classification accuracy on the Hepatitis data set than the others with applied and non-applied filter-based feature selection. Moreover, we declare that the best filter-based feature selection is Principal Components because of the highest classification accuracy obtained with for hepatitis patients.     
Keywords : Hepatitis, Machine Learning, Feature Selection, Classification, Diagnosis