How to use PCA step by step for Dimension Reduction?

I've more than 1000 images and I've extracted their features. I want to perform PCA but I'm new to Matlab. Can someone please guide the steps for Dimension Reduction on the Extracted Feature Vectors? I saw matlab fucntions like pca(), princomp(), pcares()but I don't know in which sequence to use these. I directly used the function pca() on extracted features and then trained SVM but the results were discouraging. Thanks in Advance!!!

Answer 1

[Ahmad Abdullah]

Can someone give a quick input that why PCA() returned coeff does not show good results in dimension reduction ?