News for hackers. Stuff that matters.
Re:Machine Learning books suggested by Michael I. Jordan fro... by dagw on Thursday April 06, 2017 @07:30AM (#14049606)
Essential perhaps if you want to do research in Machine learning or work on developing new machine learning algorithms and libraries. Hardly essential if all you want to do is take a well understood algorithm from a well known ML package and apply it your data. What is essential then is knowing the relative strength and weaknesses of the different existing approaches and knowing which one to pick given your data and computational limitations. And as far as I can tell non of those books cover that.

That being said, the list is excellent for people who want a solid theoretical grounding of the underlying mathematics, and many of my favorite books are on the list.

Re:Machine Learning books suggested by Michael I. Jordan fro... by Kurtz79 on Thursday April 06, 2017 @07:34AM (#14049626)
So, any resource suggestions (not necessarily books) for someone interested in a more practical approach ?

Edit: Thanks!

Re:Machine Learning books suggested by Michael I. Jordan fro... by dcl on Thursday April 06, 2017 @08:19AM (#14049862)
I recommend Elements of Statistical Learning, free @ https://statweb.stanford.edu/~tibs/ElemStatLearn/

This book covers everything from simple regression and classification from the statistical side to things like gradient boosted decision trees and the like on the ML side with enough math to make sure you understand what's actually going on.

I should note, it doesn't touch deep learning, which is what I suspect most people interested in 'machine learning' without any background in stats want to learn about these days.

Re:Machine Learning books suggested by Michael I. Jordan fro... by dzenos on Thursday April 06, 2017 @07:56AM (#14049740)
I would suggest "Hands-on Machine Learning with Scikit-Learn and TensorFlow" if you search for a practical approach.

You can find accompanying code here: https://github.com/ageron/handson-ml