8
$\begingroup$

Elements of Statistical Learning (ESL) is a book that has fantastic breadth and depth. It covers the essentials to the very modern methods by citing the papers where these original studies come about. However, I really find the language of the book very very prohibitive. I believe there is an easier way to discuss concepts. I find ESL simply too overwhelming. Can someone suggest alternatives that are friendlier to the uninitiated?

I found the sibling to ESL: Introduction to Statistical Learning. That is tone I want to read and understand. It is accommodating, without dumbing things down. Any thing similar to Intro to SL?

$\endgroup$
5
  • 2
    $\begingroup$ Can you say more about what you are looking for that you are not finding in Introduction to Statistical Learning? $\endgroup$ – Matthew Drury May 31 '15 at 3:37
  • $\begingroup$ There are certain sections in ESL that are not found in the introduction. Perhaps because it is 'beyond' being an introduction. For example, sections that talk about Reduced rank regression are not mentioned in Intro, but are discussed at length in ESL. However, (my impression is that) the writing in ESL is done in a way that burdens the reader instead of inspiring him. Of course that is just my opinion and may not be true to other readers. $\endgroup$ – cgo May 31 '15 at 5:37
  • $\begingroup$ I also noticed that in Chapter 3, ESL jumps from single output systems to multiple output systems and again to single output. It is quite confusing. And if you are already lost in the middle, then reading succeeding sections is just not productive. This may as well be a letter I should write to the authors. $\endgroup$ – cgo May 31 '15 at 5:38
  • 2
    $\begingroup$ You could try Applied Predictive Modeling by Kuhn et al. The overlap could be considerable though. $\endgroup$ – spdrnl Jun 1 '15 at 12:16
  • $\begingroup$ Is "Introduction to Statistical Learning with R" too elementary ? By basically the same authors. $\endgroup$ – meh Jun 10 '19 at 14:25
3
$\begingroup$

I agree that An Intro to Statistical Learning has a very accommodating tone. You may want to look at Learning From Data, A Short Course by Yaser Abu-Mostafa et al. I found this book and the accompanying youtube videos to be great.

Lastly, spdrnl's comment about Applied Predictive Modeling by Kuhn is a good suggestion. I have not read it yet, but I have perused it and it seems like a great resource as well.

$\endgroup$
3
$\begingroup$

Possible alternatives:

  • Pattern Recognition and Machine Learning by Christopher Bishop: I don't like the book's notation systems, but I heard the graphical model chapter is good

  • Machine Learning: A Probabilistic Perspective by Kevin P. Murphy: like a dictionary, describe various of pre-deep-learning-era machine learning methods

  • Deep Learning Book: Newer, covering more about deep learning part

  • Dive into Deep Learning: Possibly newest deep learning book so far

Also, try some course notes:

$\endgroup$

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.