Based on this post, I want to digest Elements of Statistical Learning. Fortunately it is available for free and I started reading it.

I don't have enough knowledge to understand it. Can you recommend a book that is a better introduction to the topics in the book? Hopefully something that will give me the knowledge needed to understand it?


Is a strong background in maths a total requisite for ML?

  • 12
    $\begingroup$ I found Strang's Linear Algebra and its applications to be extremely useful in understanding the matrix manipulations which form a large part of the elements. $\endgroup$ – richiemorrisroe Nov 26 '11 at 9:13

I bought, but have not yet read,

S. Marsland, Machine Learning: An Algorithmic Perspective, Chapman & Hall, 2009.

However, the reviews are favorable and state that it is more suitable for beginners than other ML books that have more depth. Flipping through the pages, it looks to me to be good for me because I have little math background.

  • $\begingroup$ Looks great - very accessible. $\endgroup$ – B Seven Nov 26 '11 at 5:38
  • $\begingroup$ I downloaded and read the "sample" - all 19 pages (wow). It is much easier to understand than The Elements of Statistical Learning. Definitely seems to be what I am looking for. Thanks. $\endgroup$ – B Seven Nov 26 '11 at 6:11
  • 4
    $\begingroup$ I have made edits to your question to provide a citation for the book. Generally speaking, putting things like "I like this one" in an answer is discouraged since if the link breaks, no one will know what "this one" was referring to. Cheers. $\endgroup$ – cardinal Nov 26 '11 at 16:47
  • $\begingroup$ I just got this and started reading it (first 75 pages). It is awesome. Very easy to understand, yet is detailed enough to be practical and useful. Highly recommended for anyone who wants to use Machine Learning. Exactly what I was looking for. Thanks! $\endgroup$ – B Seven Dec 4 '11 at 21:00

The authors of Elements of Statistical Learning have come out with a new book (Aug 2013) aimed at users without heavy math backgrounds. An Introduction to Statistical Learning: with Applications in R

The free PDF version of this book can currently be found here.

  • $\begingroup$ I was going to suggest this since its a recent release and is obviously strongly related to the objective text of the poster. Good recommendation. $\endgroup$ – Chris Simokat Sep 10 '13 at 23:08
  • 3
    $\begingroup$ Better still, the authors have announced that a free online pdf of this book will be available from January 2013 (it is being used in a MOOC which they are running.) $\endgroup$ – Flounderer Dec 19 '13 at 21:28

I found Programming Collective Intelligence the easiest book for beginners, since the author Toby Segaran is is focused on allowing the median software developer to get his/her hands dirty with data hacking as fast as possible.

Typical chapter: The data problem is clearly described, followed by a rough explanation how the algorithm works and finally shows how to create some insights with just a few lines of code.

The usage of python allows one to understand everything rather fast (you do not need to know python, seriously, I did not know it before, too). DONT think that this book is only focused on creating recommender system. It also deals with text mining / spam filtering / optimization / clustering / validation etc. and hence gives you a neat overview over the basic tools of every data miner.

Chapter 10 even deals with stock market data, but the focus is not on time series data mining. Maybe the only drawback (for you) of this excellent book.

  • $\begingroup$ It's available on Safari Books Online safaribooksonline.com. Thanks. $\endgroup$ – B Seven Nov 26 '11 at 14:07
  • 1
    $\begingroup$ Got this book and started working through it. It is very practical. In the first 18 pages you implement a complete (basic) recommendation engine. $\endgroup$ – B Seven Nov 26 '11 at 15:41
  • $\begingroup$ Wow, this book is really incredible. It teaches you how to implement all sorts of Machine Learning algorithms with just a little Python code. One of the most practical books ever. The only drawback is that Python has been updated since the book was published. It also uses many API's which have also changed. So I don't think the examples will work without some tweaking. $\endgroup$ – B Seven Dec 4 '11 at 21:17
  • $\begingroup$ @BSeven thank you, did not know that. I am not sure whether I prefer a book which utilizes pre-existing libraries (which is generally a could thing) or its own code (which works for all book examples but may be less robust due to less users). $\endgroup$ – steffen Dec 5 '11 at 10:15
  • 1
    $\begingroup$ I think these days the only choice is pre-existing libraries. They are ubiquitous, easy to integrate, cross-platform, multi-language, and fast. Besides that, if a book has its own code, it is much more difficult to modify. It's easier to modify calls to a library. Thanks for the recommendation. It is a great resource. $\endgroup$ – B Seven Dec 5 '11 at 14:08

Introduction to Machine Learning, by E. Alpaydin (MIT Press, 2010, 2nd ed.), covers a lot of topics with nice illustrations (much like Bishop's Pattern Recognition and Machine Learning).

In addition, Andrew W. Moore has some nice tutorials on Statistical Data Mining.

  • $\begingroup$ (+1) Do not know the book, but the tutorials of Andrew Moore are great (and even entertaining sometimes) $\endgroup$ – steffen Nov 26 '11 at 11:55
  • $\begingroup$ @steffen I'd recommend Radford Neale's Statistical Methods for Machine Learning and Data Mining too. $\endgroup$ – chl Nov 26 '11 at 12:11
  • 1
    $\begingroup$ +1 Alpaydin is the right way to go. I was in the exact same situation as the OP a few months ago. Struggling badly with Tibshirani, and then came across Alpaydin and things have been much better since. Eventually though I think Tibshirani is a must read. $\endgroup$ – Andy Nov 26 '11 at 20:47

Mayhaps Wasserman's All of Statistics would be of interest. You can sample the book from the link given - and just the first few paragraphs of the preface make a hard sale to your market - and you can likely download the book free through Springer if you are associated with a university.

EDIT: Oops, didn't notice how ancient this thread was.

  • 5
    $\begingroup$ Doesn't matter, the recomendation is still useful for everyone else who reads the thread (like me ;o). $\endgroup$ – Dikran Marsupial Oct 23 '12 at 16:56
  • 1
    $\begingroup$ Great book, but in fairness, if one can read and comprehend All of Statistics, a good portion of ESL is redundant. $\endgroup$ – usεr11852 says Reinstate Monic Dec 6 '15 at 7:03

The Elements Of Statistical Learning might be a tough read, especially for a self-learner. While searching for some explanations on the second chapter I have stumbled on the following resource: https://waxworksmath.com/Authors/G_M/Hastie/WriteUp/Weatherwax_Epstein_Hastie_Solution_Manual.pdf. It contains 100+ pages of annotations and explanations that clarify some complicated moments of the book. A great resource for everyone reading this book. This complementary text includes solutions for exercises.


I'd strongly recommend A First Course in Machine Learning by Rogers and Girolami. It covers the key ideas in a very logical order, with good examples and with the minimum level of maths to have a proper grounding in the fundamentals. It doesn't have the breadth of coverage of some books, but that is exactly why it is so good as an introductory text.

  • $\begingroup$ Looks like a good first book. And, there is a Kindle version. $\endgroup$ – B Seven Sep 10 '13 at 22:58

Another book that is very interesting is Bayesian Reasoning and Machine Learning by David Barber. The book is available as a free download from the author's website: http://www.cs.ucl.ac.uk/staff/d.barber/brml/


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.