Progress Update (September 2024)
Three new chapters!
We’ve inserted materials on dealing with missing data into the fourth chapter slot, bumping subsequent chapters up a number. That finished out the “Preparation” part of the materials.
Additionally, there are two new chapters:
- Measuring Performance with Resampling is fairly comprehensive. It discusses what resampling is (and isn’t), specific types of resampling schemes, and different statistical estimates.
- Grid Search discusses different grid types, especially space-filling designs. It also includes materials on parallel processing, submodel tricks, racing, and optimization bias/nested resampling.
We’ve also created the corresponding chapters in the tidymodels computing companion. There are a few gaps to fill in there, but the tidymodels group has working (but unfinalized) code/APIs that will be in their respective packages soon.
Next on deck for the “Optimization” part are chapters on iterative search and one on comparing models.
We will probably include a chapter on removing predictors. This won’t be extensive (it could be an entire book) but will give readers a sense of “I’m in danger” along with references for more information.