Vocabulary of concepts and terms in chemometrics (IUPAC Recommendations 2016)

Abstract

Abstract:Recommendations are given concerning the terminology relating to chemometrics. Building on ISO definitions of terms for basic concepts in statistics the vocabulary is concerned with mainstream chemometric methods. Where methods are used widely in science, definitions are given that are most useful to chemical applications. Vocabularies are given for general data processing, experimental design, classification, calibration and general multivariate methods.

PLS, iPLS, GA‐PLS models for soluble solids content, pH and acidity determination in intact dovyalis fruit using near‐infrared spectroscopy

“…Various spectra pre‐processes were applied, such as smoothing, first and second derivatives of Savitzky–Golay and multiplicative scatter correction (MSC), all by varying windows at 3, 11, 21, 31, 41, 51, 61 and 71 points. Smoothing was used to remove noise from the spectra by maintaining its original shape . The first and second derivatives of Savitzky–Golay were applied to correct the baseline and to improve resolution of the spectral bands .…”

Spectroscopy with computational analysis in virological studies: A decade (2006–2016)

“…Spectral normalization techniques are used when it is necessary to remove spectral changes responsible for the thickness or concentration of the sample, making the normalized spectra become comparable to each other [23]. Among the possible normalizations, there is the min-max normalization, which can be applied when there is a known peak that is stable and consistent between the specimens [23]; or scaling methods to equalize the importance of each variable in multivariate data [46]. In biological samples, the amide I (~1650 cm À1 ) or amide II (~1550 cm À1 ) peak normalization [33,34] are typically used; or vector normalization, where each spectrum is divided by its Euclidean norm (appropriate normalization after using differentiation as pre-processing) [33,34].…”

“…Multivariate analysis techniques are employed to analyze multivariate data, meaning data having two or more variables per object [46]. Examples are first-order data (such as FTIR, NIR, Raman spectrum) and second-order data (such as EEM fluorescence).…”

“…PLS is a multivariate calibration technique that finds factors (latent variables, LVs) in the spectra set that explain the maximum variance in the reference variables set, using the simultaneous decomposition of the two [46]. For this, PLS finds a set of new maximally correlated variables orthogonal to each other, similar to PCA.…”

“…For discriminatory purpose, partial least squares discriminant analysis (PLS-DA) is employed. PLS-DA is a linear classification technique for which the classification criterion is obtained by PLS analysis [46]. Therefore, it makes use of PLS to find a straight line that divides the data space into two-regions, where each region is related to the space of each class [57].…”