Publishing Is Powerful as It Serves as a Catalyst for Scope and Writing Decisions
An interview with Christoph Molnar, author of the book- Interpretable Machine Learning
A series of interviews highlighting the incredible work of writers in the space of data science and their path of writing.
"If you don’t see the book you want on the shelf, write it" – Beverly Cleary
In an endeavor to bring such notable work to the forefront, I started an interview series last year. During the first season, I presented stories from established data scientists and Kaggle Grandmasters, who shared their journey, inspirations, and accomplishments. For the second season, I’m interviewing book authors. As a writer myself, I have tremendous respect for people who write books. A single well-written article takes a lot of time, energy, and patience, and to replicate the same for a book is no mean feat. As such, this edition of the interviews will bring to light the story of some of the well-known authors in the data science field.
Meet the Author: Christoph Molnar
Christoph Molnar is a Machine Learning expert and an independent author with a statistics background. He obtained his bachelor’s and master’s degree in Munich and is currently pursuing his Ph.D. in interpretable machine learning. Previously, Christoph worked as Data Scientist at a start-up in the financial industry, where he worked on developing machine learning models. Later, he took a more traditional role as a statistician at a registry for patients with rheumatic diseases, where he worked with rheumatologists to study the effects of certain drugs.
If you have ever wanted to start in interpretable machine learning, the first book you might have picked would be Christoph’s Interpretable Machine Learning. The clarity in language, realistic examples, and the art to break down complex theories into more straightforward and understandable bits are a few of the many USPs of the book. A second edition of the book has been recently released, and if you want to lay your hands on it, you can find all information here.
Q: How did the idea of this book originate?
Christoph: In Zurich, I worked part-time and spent each Friday learning new things about machine learning. Initially, I took a course on Deep Learning. Then, I started reading research papers. I managed to lay my hands on the famous LIME paper, which presented a method to interpret predictions for black-box machine learning models. Because of my background in statistics, I always thought machine learning lacked interpretability, so I really liked the idea of the paper.
I was motivated to learn more and started looking for other interpretation approaches like blog posts and books. However, I didn’t find much material about machine learning explainability. I decided to read papers and summarize the methods in a book. This became my new "Friday" project. I published an in-progress version of the book online for free and got tremendous feedback that encouraged me to turn it into a comprehensive book.
Q: Could you summarize the main points covered in the book for the readers?
Christoph: Interpretable Machine Learning covers various methods for interpreting machine learning models. Each chapter covers one way of interpreting models. I wanted the book to be relevant for a long time, so I mostly covered so-called model-agnostic methods. These apply to any machine learning model. Some notable examples of such techniques are – permutation feature importance, Shapley values, LIME, Accumulated Effect Plots, etc. But since neural networks are widely used everywhere, a large section also covers interpretation methods that specifically address deep learning.
Q: Who do you think is the target audience for the book?
Christoph: Interpretable Machine Learning is for everyone who wants to learn how to explain machine learning models. I know that many data scientists who build predictive models are readers of the book. Then there are students and teachers of machine learning, technical managers, and many other people.
Q: What, according to you, is the best way to make the most out of this book?
Christoph: I used to think my readers would just read jump from chapter to chapter, based on which method they want to learn. Like a reference type of book. I was surprised to learn that many people still read it from cover to cover.
If you are new to machine learning interpretability, reading Interpretable Machine Learning will give you an excellent overview of topics. If you are currently working on a predictive model, you can use the book as a reference. For example, the book will allow you to choose the interpretation method that makes sense for your problem. But you can also review the limitations and issues with each interpretation method.
Q: What advice would you give a new writer, someone just starting?
Christoph: Publish early and often. Publishing is powerful. It serves as a catalyst for scope and writing decisions, helps get feedback from early readers, is good marketing, and makes the writing process more of a conversation with the readers.
Q: Who is your favorite author (in technical or non-technical space)?
Christoph: I love the fantasy books by Brandon Sanderson.
👉 Are you looking forward to connecting with Christoph? Follow him on Twitter.
👉 Read other interviews in this series:
Don’t just take notes – turn them into articles and share them with others
You do not become better by employing fancy techniques but by working on the fundamentals
Share This Article
Towards Data Science is a community publication. Submit your insights to reach our global audience and earn through the TDS Author Payment Program.
Write for TDS
