📚 Some people may know that I host a DS/ML book club where we read one book per month. This month, we read ‘Modeling Mindsets’ by Christoph Molnar. Christoph is a statistician and machine learning expert. I have read his first book Interpretable Machine Learning and absolutely loved it.
This Modeling Mindsets book elucidates the worldviews behind various statistical modeling and machine learning mindsets, such as Bayesian inference, supervised learning, causal inference, and more. It’s a short and pleasant read.
In this blog post, I’d like to share an interview we had with the author that’s full of insights. I’ll also be sharing my personal reading notes for those who are curious about this book.
🔗 Book link: https://amzn.to/3PVDto5
Our DS/ML book club had a great chat with Christoph about this book. Here are some key points from our discussion:
- Frequentism vs. Bayesianism vs. Likelihoodism
- The T-Shaped Modeler
- The LLM mindset
- The challenges in interpreting LLMs
- Career advice
Chapter 1 Introduction
- Model: a mathematical model that consists of variables and functions. Variables have different names in different mindsets. Functions related the variables to each other. Modelers use data to find the best function to relate the variables.
- The purpose of the model — how to use and interpret it — depends on the modeling mindset. A modeling mindset provides the framework for modeling the world with data.
Chapter 2 Statistical Modeling
- The data are generated by a process that can be approximated by relating variables and specifying their distribution. Use statistical models to reason under uncertainty.
- Modelers fit models by…