Workflowsets in {tidymodels} provide a useful way for analysts to run several models, even tune several machine learning models on their dataset, simultaneously. You can then extract model fit information, predictions, and make comparisons to identify the most effective model type and the optimized tuning parameters.
Below are links to two workflowsets tutorials I’ve written. The first link is to an older tutorial that wrote and posted in the blog 3 months ago. This tutorial builds several machine learning models on NWSL data to solve a classification task. The second link is a more recent tutorial that walks through workflowsets and model tuning for outcomes that are both binary and continuous.
Both tutorials go over model building, hyperparameter tuning, model comparisons, making predictions, and storing the final model for deployment at a later time in {tidymodels} using workflowsets.
Hopefully both tutorials provide analysts with a clear overview of how to set up workflowsets and run several models in parallel, helping to make their work more efficient.
