[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
on the "uniqueness" of Singular value decomposition (answer to the blog question)
- To: Rao Kambhampati <rao@asu.edu>
- Subject: on the "uniqueness" of Singular value decomposition (answer to the blog question)
- From: Subbarao Kambhampati <rao@asu.edu>
- Date: Thu, 25 Feb 2010 07:01:29 -0700
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:sender:received:date :x-google-sender-auth:message-id:subject:from:to:content-type; bh=3RKHUvIrCtLexZy0qHd/HN9Myo8HOVNAW9rLe5I1FEA=; b=RQ39D7tHVqtPlPsgDm5zQym7J0CHk3V7wKvYayvEpR3mp760tVxt/HSsDRsVIL2e62 JuV2SP/kdZ2FtWKc1VSbONfhrdn2XAb9f8JxWPoxDEGgrD68YXIjgMwqdiMgw05Zirld kcSc6FoZtcSlnPZwZAdrKa4LS2BfRUDt5Vf9c=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:sender:date:x-google-sender-auth:message-id:subject :from:to:content-type; b=nzaVcXrmwzuhESbGY66/9TF2IMiTRjstKjIFQjvbRQ4vnGlaZHu8dY25SU+oUA+12u dpBq9+O/d+IbSY5bqr6wfZUtmffSS66a0NtJS4Ju87XrPkwr/91X1hvbp3v5GAoPHt/9 o6QKbAYTyKKITdQyWNRlWOBie9Ie3a5dmwCJI=
- Sender: subbarao2z2@gmail.com
There is a question on the blog as to whether SVD is guaranteed to be unique for a given matrix.
dt= df * ff * tf'
It is easy to see that ff is unique--since it is just the square roots of eigen values of dt*dt' (or dt'*dt).
The question is about df and tf' . Their "uniqueness" depends on whether or not the singular values are all distinct (which in turn
depends on whether the eigen values of dt*dt' are distinct)
(Recall that a square matrix can have repeated eigen values--which happens when its characteristic equation has a factor--say
(x - l)^k --so we have k roots at l, and the eigen value l is said to have multiplicity k )
When the singular values are all distinct, then df and tf' are unique *upto* the sign of the vectors (that is, their magnitude is unique, but sign can be either positive or negative--but the signs of df and tf' have to be consistent).
When the singular values are degenerate, then df and tf' are no longer unique.
The lack of uniqueness doesn't however affect any of the low-rank approximation properties of SVD we cared about in doing LSI.
Rao
ps: Here is a longer technical description on SVD uniqueness: http://informatik.unibas.ch/lehre/hs09/cs253/slides/svd09.pdf