We’ve been working on developing machine learning models using {tidymodels} over the past several weeks. Sometimes, though, you need to build a variety of models on your data. This week, Ellis Hughes and I explore {workflowsets}.
Workflow sets allow you to set specific recipes for different model types and then run a variety of models simultaneously. Topics we cover:
- Setting up recipes
- Specifying multiple models
- Creating a workflowset
- Fitting all models within the workflowset
- Choosing the optimal model
- Two different approaches to fitting the optimal model on your test data set
To watch our screen cast, CLICK HERE.
To access the code and data (courtesy of kaggle), CLICK HERE.
If you like what we are doing and would like to buy us a coffee or beer, visit our Patreon page HERE.