New classical least-squares/partial least-squares hybrid algorithm for spectral analyses

A new classical least-squares/partial least-squares (CLS/PLS) hybrid algori thm has been developed that demonstrates the best features of both the CLS and PLS algorithms during the analysis of spectroscopic data. By adding our recently reported prediction-augmented classical least-squares (PACLS) to the hybrid algorithm, we have the additional benefit that known or empirica lly derived spectral shape information can be incorporated into the hybrid algorithm to correct for the presence of unmodeled sources of spectral vari ation. A detailed step-by-step description of the new hybrid algorithm in c alibration and prediction is presented. The powerful capabilities of the ne w PACLS/PLS hybrid are demonstrated for near-infrared spectra of dilute aqu eous solutions containing the analytes urea, creatinine, and NaCl. The PACL S/PLS method is demonstrated to correct the detrimental effects of unmodele d solution temperature changes and spectrometer drift in the multivariate s pectral calibration models. Initially, PLS and PACLS/PLS predictions of ana lytes from variable-temperature solution spectra were made with models base d upon spectra previously taken of the samples at constant temperature. The presence of unmodeled temperature variations and system drift caused the p rediction errors from these models to be inflated by more than an order of magnitude relative to the cross-validated errors from the calibrations. PLS achieved improved predictions of the variable-temperature spectra by addin g spectra of a few variable-temperature samples into the original calibrati on data followed by recalibration. PACLS/PLS predictions were corrected for temperature variations and system drift by adding spectral differences of the same subset of samples collected under constant- and variable-temperatu re conditions to the PACLS prediction portion of the hybrid algorithm durin g either calibration or prediction. Comparisons of the prediction ability o f the hybrid algorithm relative to the PLS method using the same calibratio n and subset information demonstrated hybrid prediction improvements that w ere significant at least at the 0.01 level for all three analytes. The new hybrid algorithm has widespread uses, some of which are also discussed in t he paper.