A new family of genetic algorithms for wavelength interval selection in multivariate analytical spectroscopy

Abstract

A new procedure is presented for wavelength interval selection with a genetic algorithm in order to improve the predictive ability of partial least squares multivariate calibration. It involves separately labelling each of the selected sensor ranges with an appropriate inclusion ranking. The new approach intends to alleviate overfitting without the need of preparing an independent monitoring sample set. A theoretical example is worked out in order to compare the performance of the new approach with previous implementations of genetic algorithms. Two experimental data sets are also studied: target parameters are the concentration of glucuronic acid in complex mixtures studied by Fourier transform mid-infrared spectroscopy and the octane number in gasolines monitored by nearinfrared spectroscopy.

Enhanced genetic operators design for waveband selection in multivariate calibration based on NIR spectroscopy

“…, Goicoechea et al . , and Leardi et al . employed them for variable selection, grouped in intervals in .…”

“…, and Leardi et al . employed them for variable selection, grouped in intervals in . In the chemometrics field, not as many approaches as could be desirable have been developed related to a specific design of parts of the GA.…”

“…Some examples are those of Goicoechea et al . , where they do not use a separate validation set and design a specific initialization strategy for selecting both wavelengths and wavebands; of Lavine et al . , where they examine, by means of a GA, plots from two or three principal components in order to determine subsets of variables, not necessary bands; and of Fei et al .…”

Sequential application of backward interval partial least squares and genetic algorithms for the selection of relevant spectral regions

“…It should be noted that the main purpose of this study is to merge and obtain synergy between two different methods for spectral region selection and that several other methods for region selection have been developed and described in the literature [6][7][8][9].…”

Partial Least Squares Discriminant Analysis Model Based on Variable Selection Applied to Identify the Adulterated Olive Oil

“…Moreover, the problem of singular values that is caused by the number of variables much greater than the number of samples can be avoided in some multivariate analysis methodologies. In recent years, with the development of chemometrics, a large number of variable selection methods have been proposed including genetic algorithms (GAs) (Goicoechea and Olivieri 2003;Boschetti and Olivieri 2004;Leardi and Nørgaard 2004), uninformative variable elimination (UVE) (Centner and Massart 1996), the competitive adaptive reweighted sampling method (CARS) (Li et al 2009), successive projection algorithm (SPA) (Paiva et al 2012;Araújo et al 2001), variable importance in projection (VIP) (Chong and C. H. Jun. 2005), and Canonical Measure of Correlation (CMC) index (Ballabio et al 2010).…”