Part 3 is Finished, Part 4 Started
updates
tidymodels
We’ve released additional chapters in the last month or so. These conclude Part 3 of the book.
- Iterative Search shows how to optimize tuning parameters using methods such as simulated annealing, genetic algorithms, and Bayesian optimization. This chapter also lays some of the groundwork for general Bayesian methods.
- Feature Selection is a survey of different methods and issues regarding the judicious removal of predictors.
- Comparing Models has a statistical focus on different ways to compare predictive performance within or between models.
On to Part 4 (Classification). So far, we have
- Characterizing Classification Models is a pretty comprehensive discussion about the myriad tools for determining whether your classifier is doing what you want it to do.
- Generalized Linear and Additive Classifiers has also been started but only contains some exploratory data analysis for the example that is used in three chapters (the forestation data). Thanks to Simon Couch for reviewing that section and generating the data.
For these chapters, we may not write them as linearly as the previous set.
We’ve also updated the tidymodels computing supplement for these chapters.
One nice feature of these updates is that many sections of the book have direct links to the computing supplement for that topic (as well as a return link).