Automatic Hyperparameter Optimization of Bagging Workflows

Authored by: Fábio Pinto and Vítor Cerqueira

autoBagging is an R package for automatically optimizing bagging workflows to solve classification predictive tasks.


Currently, autoBagging is only available in Github. Soon it will be submitted to CRAN.

Install the package using devtools:

followed by:

In some OS, the installation might need manual installation of recursive dependencies (e.g. data.table).


The core function is autoBagging. Its input is simply the formula for the predictive classification task and the dataset:

For predicting new instances, the model uses the standard predict method:

Contact us at: {fhpinto, vmac}