12
$\begingroup$

I am trying to learn Reduced-Rank Regression (RRR) from The Elements of Statistical Learning. I find the writing and them mathematics a little too prohibitive. Does any of you have a resource/text/introduction/tutorial that is friendlier as an introduction?

For example, I found a text on Canonical Correlation Analysis (CCA) from Penn State an excellent layman's introduction to the topic. We can move on to the mathematics after being able to get the intuition.

$\endgroup$
1
  • 3
    $\begingroup$ I now came across these two questions of yours about RRR and noticed that they have only a single upvote each, and it's mine. This confirms my impression that there are very very few people on this forum interested in RRR. $\endgroup$
    – amoeba
    Oct 30, 2015 at 17:26

1 Answer 1

12
$\begingroup$

I agree that Section 3.7 in The Elements of Statistical Learning is quite confusing. I love this book but hate this section. The PDF available online has this nice Scream pictogram near the section title, so don't worry if you find the math there too complicated.

Section 3.7

Here are some other references. There is one book specifically focusing on RRR:

And there is a textbook on multivariate statistics with good coverage of RRR:

Note that this is the same Izenman who has originally coined the term "reduced-rank regression" in

I don't know of a good tutorial paper, but have recently come across this PhD dissertation (that is essentially a composition of three separate papers, available elsewhere too):

Most of it is quite technical, but it can be useful to read the introduction and the beginning of the first main chapter. There is also lots of further references there, in the literature review in the beginning.


I am only giving references here, but I might post a more substantial answer about RRR in your other question What is "reduced-rank regression" all about? when I have some spare time and if there is any interest.

$\endgroup$
3
  • 2
    $\begingroup$ Sure...I would be interested in that post...isn't RRR basically dimension reducing? $\endgroup$ Oct 30, 2015 at 17:29
  • 2
    $\begingroup$ Good to know, @DJohnson. By the way, one way of expressing interest here is to upvote. If you find that (or this) question interesting, please upvote it. If you find this answer helpful and even want to hear more, please upvote it. I am only writing this because I can see that you are relatively new here and are being quite active (which is great!), but don't seem to be upvoting almost at all (23 votes for 580 rep). Have you only seen 23 posts here that have been of any interest to you? You have even written 59 answers yourself! Hopefully most of those questions were interesting to you! $\endgroup$
    – amoeba
    Oct 30, 2015 at 18:50
  • 2
    $\begingroup$ Thanks for the reminder re upvoting. Your numbers are a reminder that I should be more liberal with this CV option. $\endgroup$ Oct 30, 2015 at 18:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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