stacks |
An R package for model stacking that aligns with the tidymodels. It uses a regularized linear model to combine predictions from ensemble members. |
CRAN and GitHub |
Super Learner |
No |
SuperLearner |
An R package that provides the original Super Learner and includes a clean interface to 30+ algorithms. |
CRAN and GitHub |
Super Learner |
No |
subsemble |
An R package that provides stacking via the super learner algorithm and also implements the subsemble algorithm. |
GitHub only |
Super Learner and Subsemble |
Yes |
caretEnsemble |
An R package that provides an approach for stacking, but it implements a bootstrapped version of stacking. |
CRAN and GitHub |
Bootstrap Stacking |
No |
h2o |
An R package that provides an efficient implementation of stacking and allows you to stack existing base learners, stack a grid search, and also implements an automated machine learning search with stacked results. |
CRAN and GitHub |
Stacked Ensembles and AutoML |
Yes |