Comments on: A List of Data Science and Machine Learning Resources

By: Nate Weiss

Nate Weiss — Fri, 01 Feb 2013 20:49:05 +0000

Thanks Karan! Yeah MJ really breaks it down very clearly via a decision theoretic approach. Plus it is nice that it is without any of the hype that is out there right now wrt Bayes/Frequentest.
Check out the next post in the series – http://conductrics.com/data-science-resources-2 . I have a few links on clustering and RL, plus a shout out to Stephen Boyd and his convex optimization course. His class is kinda awesome actually, esp if you already have some ML behind you, since he approaches it from a different angle.

By: Karan Sikka

Karan Sikka — Fri, 01 Feb 2013 20:29:55 +0000

Really good links on ML. Keep up the good work. I really liked your links on MJ lecture since it is uncommon to find lectures on fundamental topics like frequentist vs Bayesian

By: Ben

Ben — Thu, 27 Dec 2012 20:26:21 +0000

Hello Matt,

thanks for your collection. I have a question: I’m a physics student and due to my research topics (stochastic processes) I want to learn machine learning. I know how to programm, but I have never done anything ML-related before. There are also no classes/lectures for ML at my university, so I feel a little bit lost. You have posted great material, but most of it is video material. Could you give me a guide-line where I should start for the very basics? Is there THE book you recommend on machine learning?

Looking forward to get a reply. Best greetings,
Ben

By: Ivan Savov

Ivan Savov — Tue, 18 Dec 2012 23:23:02 +0000

I forgot to say, on the Latent Dirichlet Allocation topic –> there is an excellent introductory paper by G. Heinrich:

http://www.arbylon.net/publications/text-est2.pdf

By: Ivan Savov

Ivan Savov — Tue, 18 Dec 2012 23:20:49 +0000

Hi Matt,

Thx for the links. I totally agree with you that linear algebra is essential to all things machine learning.

I working on a book that explains all of LA in a simple and intuitive way — in just 60 pages. Check it out, and feel free to send anyone who is lacking in LA skills my way:
minireference.com/linear_algebra/introduction (free for now while fixing typos, available in print in Jan)

I would love to hear what you think.

Peace out,

Ivan

By: LeeRit

LeeRit — Mon, 17 Dec 2012 08:23:04 +0000

Thanks for sharing the resources. I’ve bookmarked the page.

I’m not so sure about other areas but for Statistical Machine Learning, I completely agree that Philip Koehn is definitely the guy to go for. I’m using his Moses package as well.