Kidney Disease Diagnosis using Classification Algorithm

Abstract

Kidney disease is a larger problem in today’s era. Huge population suffers due to kidney problem and mortality rate is increasing largely. Diagnosis of kidney disease and dropping death rate in population is challenging task for health care officials. In this study, we propose the methodology to diagnosis kidney disease using Naive Bayes (NB), Random Forest, Bagging and k-Nearest Neighbor (KNN). Missing value in dataset is handled by Mean imputation technique and transformation is done using LabelEncoder method to transform nominal dataset to numeric dataset. Accuracy percentage and Root Mean Square errors were calculated and compared. From the obtained results Decision Tree algorithm gave an improved accuracy percentage of 100% for diagnosing kidney disease.