Crossvalidation techniques in PLSR

I have a data matrix X of size 85x401 where columns correspond to reflectance values of various soil samples from 400 to 800nm with 1nm increment and rows correspond to the observations.I have another data matrix Y of size 85x1 containing the values of organic matter content.I need to predict OM content from spectroscopic data using PLSR.

Inorder to do the first step in PLSR method, we have to choose the optimum no.of components to be used in model which is decided by cross validation techniques.

Do v have any matlab code to do crossvalidation and determine the optimum no.of components.?

Answer 1

[bym]

http://blogs.mathworks.com/loren/2011/11/21/subset-selection-and-regularization/