Effect of Features Selection Method for Kidney Disease Classification Using Naïve Bayes and J48 Classifiers

Abstract

There are many fields today that use database systems to store data.Over timethis has lead to the creation huge amount of data. The importance of these huge data is that are knowledge and relationships between them. When analyzing these data we get knowledge and facts that help decision makers to make the right decisionswhich are called data mining.The medical field is one of the most important fields to apply data mining techniques. The purpose of this study is to knowthe effect of features selection algorithms in increasing the accuracy of the classifier (model).We have applied this study on the Chronic Kidney Disease (CKD) dataset which contains 25 features used to diagnose (CKD) to know the effect of features selection algorithms in the increase classifier accuracy. In this study we have used naïve bayes and J48 classifiers with the wrapper features selection evaluator to select best features that have high effect in classifier accuracy and eliminate unimportant features through experiments we have noticed when used naïve bayes a classifier with the wrapper features selection evaluatorthe degree of accuracy increased from 95% to 99.5%. But when used the j48 classifier with wrapper features selection evaluator the degree of accuracy not significant.