Chronic Kidney Disease Detection using AdaBoosting Ensemble Method and K-Fold Cross Validation

CKD (Chronic Kidney Disease) is among the most serious health issues worldwide. Among the causes of total global mortality, chronic kidney disease ranked 27th in the Global Burden of Disease Study. Millions of people die every year as a result of an inability to afford treatment for chronic kidney disease. A kidney disease can be slowed or stopped if it is diagnosed and treated early. In order to predict CKD, machine learning is effective. In this paper, dataset from UCI repository is used with 24 features like age, sugar, albumin etc. The AdaBoosting ensemble algorithm, as well as K-Fold Cross Validation technique, is used to predict CKD. The missing data in this dataset is filled in using the Mean Imputation technique. Next, the classification of data is carried out using KNN, Decision Trees (DT), Random Forest(RF), Naive Bayes(NB), Logistic Regression(LR) and its ensembles. The results were compared with KNN, Decision Tree, Naive Bayes, Random Forest, and Logistic Regression. Experimental results shown that AdaBoost-Random Forest ensemble is 99 % accurate for early detection of kidney illness.