The 3rd Edition of Interpretable ML is published 🥳

I published the third edition of the Interpretable Machine Learning book. 🎉 🎉 🎉

Interpretable Machine Learning is the first and probably most influential book I’ve written. I started the project over 7 years ago and published the first edition in early 2019. The book got its first update 3 years later, in 2022. The second edition also coincided with my becoming a self-employed writer. But it was also the time after my PhD when I felt burned out for the topic of interpretability. You can read all about why I almost stopped working on the book in this post.

The 3rd edition was a major clean-up. I moved the book from bookdown to Quarto, improved the examples, and cleaned up the intro. I also added new chapters on Data and Models, Goals of Interpretability, Ceteris Paribus, and Leave One Feature Out Importance. So if you know the book, the third edition offers new insights. But this new edition was as much an update for the readers as it was for me: The book project is much cleaner and easier for me to work with. I got rid of a lot of technical clutter, and it feels like a weight lifted off my shoulders.

The book remains open to everyone to read. For free. It only has a new look now as I’m using Quarto to render the book’s website.

If you want to support me, or you simply prefer to read the book as an ebook or paperback, you can also get a copy: