Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.


Math Forum
»
Discussions
»
Software
»
comp.softsys.matlab
Notice: We are no longer accepting new posts, but the forums will continue to be readable.
Topic:
Question about Principal Component Analysis (PCA) results.
Replies:
1
Last Post:
Nov 6, 2012 2:25 PM



Aaronne
Posts:
110
Registered:
6/2/11


Question about Principal Component Analysis (PCA) results.
Posted:
Nov 6, 2012 8:57 AM


Currently, I am a little bit confusing about Principal Component Analysis (PCA) results and in which way I can use PCA.
For example, I have got some spectral signal of waves. They are digitised by 2,000 sampling points of each wave. And the training data set contains 500 of such waves.
I store this data set in a 2,000 by 500 matrix.
Then I run the PCA program on the data set, and I want to get 3 PCs only. My question is:
What results should I get?
I think it should produce a 2,000 by 3 matrix and each column of the result matrix is a PC. Am I right?
For example, I use Matlab based fastica package to do the PCA, A is the original 2,000 by 500 matrix,
Should I do:
numberofPC = 3;
[PCs, eigenVal] = fastica(A, 'lastEig', numberofPC, 'only', 'pca');
or
[PCs, eigenVal] = fastica(A', 'lastEig', numberofPC, 'only', 'pca');
???
Thanks very much.



