DETECTING AND ADJUSTING FOR NONLINEARITIES IN CALIBRATION OF NEAR-INFRARED DATA USING PRINCIPAL COMPONENTS

A new regression method for non-linear near-infrared spectroscopic dat a is proposed. The technique is based on a model which is linear in th e principal components and simple functions (squares and products) of them. Added variable plots are used to determine which squares and pro ducts to incorporate into the model. The regression coefficients are e stimated by a Stein estimate which shrinks towards the estimate determ ined by the first several principal components and the selected non-li near terms. The technique is not computationally intensive and is appr opriate for routine predictions of chemical concentrations. The method is tested on three data sets and in all cases gives more accurate pre dictions than does linear principal components regression.