Prediction of Chronic Kidney Disease Using Data Mining Techniques

Abstract

The kidneys have great importance in the body, as it performs many basic functions, and any defect in it harms the entire body and leads to complications and great risks that may reach death if it is not discovered and treated early, and the kidneys may be exposed to many health problems and diseases, which must be paid attention for it, therefore, it is necessary to constantly and continuously detect to predict a possible disease. There are many tools that help in predicting the presence of kidney diseases or not, but in this study we will rely on data mining techniques to predict correctly, Where this study aimed to predict kidney disease using data extraction techniques, an analytical approach was used to reach the results, was used Chronic Kidney Disease data set, and performed the preprocessing on data set, after that applied the data mining techniques by used of five techniques which are Support Vector Machine (SVM), Logistic Regression (LR), Decision Tree, K Nearest Neighbor (KNN), Naïve Bayes (NB). These techniques were applied using three tools: Weka, Orange, and Python. It concluded the SVM had a highest accuracy based on Orange it equal 99.8%.