MATLAB: How to apply The Kaiser rule in PCA

statistics

Hi,
In MATLAB there is a bulid function to apply principle component analysis PCA. However, I have a problem on applying The Kaiser rule which drop all components with eigenvalues under 1. For Example I want to apply this method on the data:
X = [1 2 3 4 5 ; -1 -3 -1 2 4 ; -2 1.5 3 2 -9 ; 1 -1 0.25 2.3 2.2];
[coeff,newdata,latend,tsd,variance] = pca(X)
Please can some one help me on this. Many thanks.

Best Answer

Hey!
Instead of calculating PCA go with SVD. Take under the under root of sigmas 's' and now you have eigenvalues. Check for kaiser rule and select the column with eigenvalue less than 1
X = [1 2 3 4 5 ; -1 -3 -1 2 4 ; -2 1.5 3 2 -9 ; 1 -1 0.25 2.3 2.2];
[u,s,v] = svd(X)
Please refer to this link for further information