Forhigh-dimensional chemometric data, the inverse matrix X t − 1 problem in regression models is a difficulty. Multicollinearity and identification result from problem. The usage of least absolute shrinkage sel...