Books For ML Theory
- Elements of Statistical Learning by Tibshirani, Hastie and Friedman is a legendary book in Machine Learning with a cult following. What strikes me as its most appealing feature is the fine balance between stating algorithms and mathematical rigor.
- An Introduction to Statistical Learning is a great place to start, especially if you haven't been trained as a mathematician. It introduces all the essential concepts in modern machine learning technology with lots of examples, graphs and pictures.
- Machine Learing: A Probabilistic Perspective by Kevin Murphy is definitely a work of art. I feel like chapter 8 is a good place to start, if you want to dive right in! A fair warning however is that this ia thick book, and it often helps to just skim through the derivations.
Online Courses
- Machine Learning by Andrew Ng is course that started Coursera, and for good reason. This is a great place to start, even if you just heard the words "Machine Learning". The only downside I can concoct is that the assignments are in GNU Octave instead of in Python.
- Machine Learning Crash Course developed by Google is probably the most modern way of teaching the subject, all the way from Linear Regression to GANs. Better yet, all coding exercises are on their playground, removing any activation energy required to install Python/TensorFlow on your computer. And there's more from Google/IBM on Coursera!
- DataCamp Career Track, without which this list would be incomplete. Over a 100 hours of intensive lectures, assignments and programming exercises in theory and on using libraries, APIs and general advice makes this career track a true legend in online learning.
