Reddit reviews Machine Learning: An Algorithmic Perspective (Chapman & Hall/Crc Machine Learning & Pattern Recognition)

Two Python ML resources below. The former mixes math with working code. The latter is new and appears to be more of a guide to applying scikit-learn and/or milk specifically.

• http://www.amazon.com/Machine-Learning-Algorithmic-Perspective-Recognition/dp/1420067184
  
• http://metarabbit.wordpress.com/2013/05/31/building-machine-learning-systems-with-python/
  
  EDIT: All of the code from the first one is here: http://www-ist.massey.ac.nz/smarsland/MLBook.html

I haven't personally read it, but the table of contents of this book looks like it has some potential to be a more modern version of what the Mitchell book aimed to provide.

I may be teaching a course in ML next year; I may order a copy to see if it's any good. Has anyone else looked at it?

Don't be. Basically everything here is either about conditional probability (naive bayes) or weights*inputs + bias > threshold calculations. The weights inputs stuff is basic neural networks. Support vectors machines, perceptrons, kernel transformations, are all about finding linearly separable classes (in some appropriate dimension).

For more info this will help http://en.m.wikipedia.org/wiki/Perceptron

If you are interested I really have enjoyed reading http://www.amazon.com/gp/aw/d/1420067184/ref=redir_mdp_mobile which explains really well a lot of machine learning stuff and demystifies the math

For some practical examples here are some blog posts I've written:

K means step by step in f#: http://onoffswitch.net/k-means-step-by-step-in-f/

Automatic fogbugz triage with naive bayes: http://onoffswitch.net/fogbugz-priority-prediction-naive-bayes/

I share the links mostly cause its nice to see a worked through, practical example with code. In the end a lot of these algorithms aren't that hard to implement with some matrix math

Anyways, hope this helps!

You can find it here on amazon.com:

http://www.amazon.com/gp/product/1420067184?ie=UTF8&tag=oliviergrisel-20&linkCode=as2&camp=1789&creative=390957&creativeASIN=1420067184

(Please feel free to strip the reference to oliviergrisel-20 in the URL should you want not to tip me through the amazon affiliates program)

I'm afraid I can't answer your question specifically, but Stephen Marsland has written a great introductory level machine learning book that has all of the example code written in python.