COMPRESSION OF FIRST-ORDER SPECTRAL DATA USING THE B-SPLINE ZERO COMPRESSION METHOD

The B-spline zero (B0) compression method has been investigated for th ree different raw spectrometric data sets, namely NIR reflectance, FT- IR reflectance and UV-VIS transmittance data The data were modelled wi th partial least squares (PLS) regression and we have investigated the compression ratio and model dimensionality versus the ability to pres erve the information of the original data as a function of knot sequen ce guide type, i.e. mean, standard deviation (SD) and relative standar d deviation (RSD) of the variables. Generally we conclude that the inv estigated data can be compressed to about 20% of the original number o f variables without loss of prediction ability or change in model dime nsionality (number of pc's) in comparison with the original data. (C) 1996 by John Wiley & Sons, Ltd.