TABLE 1.

Constituent learner functions considered and selected for use in the final Super Learner ensemble

Family	Available functions	Description	Functions selected for further consideration and inclusion in final SuperLearner ensemble
Boosting	gbm	Gradient boosting model	gbm
LASSO	biglasso	LASSO	biglasso
Elastic net	glmnet at alpha levels 0.1 – 0.9 (in 0.1 increments)	Elastic net regularized logistic regression model with varying levels of the mixing parameters between LASSO (alpha = 1) and ridge (alpha = 0)	glmnet, alpha = 0.5 glmnet, alpha = 0.8 glmnet, alpha = 0.9
K-nearest neighbor	knn	K-nearest neighbors	knn
	kernelKnn	Kernel-based K-nearest neighbors
Decision trees / regression trees	cforest	Conditional random forests	Cforest randomForest
	dbarts	Discrete Bayesian additive regression trees samples
	extraTrees	Extremely randomized trees
	ipredbagg	Bagging for classification, regression and survival trees
	randomForest	Random forest
	ranger	Fast implementation of random Forests / recursive partitioning
	rpart	Recursive partitioning and regression trees
	rpartPrune	Recursive partitioning and regression trees with pruning
Support vector machine	svm	Support vector machines	ksvm
	ksvm	Kernel-based support vector machines