Back to Resources

Great textbooks for machine learning (ML)!

Like classical statistics, ML is a set of tools that are extremely useful to have at your disposal as you try to better understand your data.

[If the statistics field had] incorporated computing methodology from its inception as a fundamental tool, as opposed to simply a convenient way to apply our existing tools, many of the other data related fields [such as ML] would not have needed to exist – they would have been part of statistics. – Jerry Friedman, taken from “Probabalistic Machine Learning” by Kevin Murphy

For many of the textbooks linked below, free versions of PDFs are likely buried on the internet!

Introductory ML

  • Applied Predictive Modeling
    • note that predictions are not just about the future, but for any value you don’t know
    • very applied, little to no math, great coverage of many topics (except not much detail on neural networks)
    • good discussions of the tradeoffs between predictive accuracy and interpretability
  • Introduction to Statistical Learning
    • thorough, easy-to-read, beginner-friendly
  • Deep Learning with Python
    • great, simple intro to neural networks
    • very practical, all code and no math
    • I found some terms/concepts weren’t clearly defined (b/c no math) and neural nets are never visually described, which I find useful
    • If you’re willing to google topics that you find insufficiently explained, this book is a great way to get started with neural networks and deep learning

Less introductory ML

  • Deep Learning
    • drink from the firehose
    • a few sections on more advanced topics were a bit hard to follow, but overall many gems here, especially the section on Structured Probabilistic Models

Foundations

My favorite ML textbook

  • Probabilistic Machine Learning
    • not finished yet, but descriptions are wonderfully clear and concise
    • while introductory, has more math/linear algebra than textbooks listed below
    • has a follow up book that covers advanced topics