0
$\begingroup$

I was going through my homework. I have a question to interpret the plot which i obtain from first and second left singular vectors after computing SVD. What can i exactly interpret, is it specific to the data set which i have or is there is any generic interpretation?. How does the plots from first and second left singular vector differ? Consider both cases i.e normalized and unnormalized(original) data.

I am describing dataset here if necessary, basically one data set contains longitude and latitude. And the other climate dataset contains the min temperatures, max temperatures, avg temperatures and rain. SVD is applied to climate data set only. The first data set is just for plotting on the map, so basically i get a heat map sort of but not sure of its interpretation. Please let me know if you need more details.

$\endgroup$
1
$\begingroup$

The singular vectors are closely related to principal components in PCA, and there is an exact correspondence if the data has unit variance and zero mean (or if you center and normalise the data before you apply SVD). I think understanding this connection will help you interpret your singular vectors. You'll find resources explaining both SVD and PCA and the connection between the two in many places on the web, such as: http://www.quora.com/What-is-an-intuitive-explanation-of-the-relation-between-PCA-and-SVD

$\endgroup$
  • $\begingroup$ What can we interpret for unnormalized data $\endgroup$ – user52705 Jun 11 '15 at 16:02

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.