Table 2.

Summary of reviewed techniques in Chronic Kidney Disease.

Ref.	Base Learner Models	Ensemble Model	Data Type	Preprocessing Technique	Positive/Negative Cases	Dataset	Attributes/Instances	Accuracy	Best Model
[45]	NB, LR, MLP, SVM, DS, RT	AdaBoost, Bagging, Voting, Stacking	Clinical	Handled missing values, feature selection, and sample filtering		Razi Hospital	42/936	Bagging = 99.1%, Boosting (AdaBoost) = 99.1%, Voting = 96.6%, Stacking = 97.1%	Boosting Bagging
[46]	NB, LR, ANN, CART, SVM	Gradient Boosting, RF	Clinical	Feature selection, handling missing values, and imputation	250/150	UCI Chronic Kidney	25/400	Bagging (RF) = 96.5%, Boosting (Gradient Boosting) = 90.4%	Bagging
[2]	-	AdaBoost, RF, ETC bagging, Gradient boosting	Clinical	Feature engineering	250/150	UCI Chronic Kidney	25/400	Bagging (Extra trees) = 98%, Bagging = 96%, Bagging (RF) = 95%, Boosting (AdaBoost) = 99%, Boosting (Gradient) = 97%	Boosting
[47]	LR, KNN, SVC	Gradient Boosting, RF	Clinical	Handled missing values	250/150	UCI Chronic Kidney	25/400	Bagging (RF) = 99%, Boosting (Gradient) = 98.7%	Bagging
[13]	-	AdaBoost, Bagging and Random Subspaces	Clinical	Feature extraction	250/150	UCI Chronic Kidney	25/400	AdaBoost = 99.25%, Bagging = 98.5%, Bagging (Random Subspace) = 99.5%	Bagging
[11]	NB, SMO, J48, RF	Bagging, AdaBoost		Feature selection and handling missing values	250/150	UCI Chronic Kidney	25/400	Bagging = 98%, Bagging (RF) = 100%, Boosting (AdaBoost) = 99%	Bagging