ABSTRACT
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794
“Local methods” is a general term for a system for quantitative analysis that searches a database of samples with known spectral and analytical data for a small number of samples that are very similar in spectral terms to a sample for which only the spectrum is known. The analysis of the unknown sample is then inferred from the analysis of the similar samples. This is a very different approach
DBURNS: “7393_c040” — 2007/7/27 — 17:17 — page 782 — #2
in near-infrared (NIR) quantitative analysis, which use a model building “calibration” step prior to the analysis of unknowns.