An integrated approach to the simultaneous selection of variables, mathematical pre-processing and calibration samples in partial least-squares multivariate calibration

• Published in: Talanta (Oxford) Vol. 115; pp. 755 - 760
• Main Authors: Allegrini, Franco, Olivieri, Alejandro C.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 15.10.2013
• Subjects: Algorithms; Calibration; Computer simulation; least squares; Least squares method; Least-Squares Analysis; Mathematical models; Monte Carlo Method; Multivariate Analysis; Multivariate calibration; Optimization; Outlier detection; Partial least-squares; Philosophy; Pre-processing selection; Sample selection; Software; Spectrum Analysis - standards; Spectrum Analysis - statistics & numerical data; Strategy; system optimization; Variable selection; Outlier detection; Pre-processing selection; Sample selection; Multivariate calibration; Variable selection; Partial least-squares
• ISSN: 0039-9140; 1873-3573; 1873-3573
• DOI: 10.1016/j.talanta.2013.06.051

A new optimization strategy for multivariate partial-least-squares (PLS) regression analysis is described. It was achieved by integrating three efficient strategies to improve PLS calibration models: (1) variable selection based on ant colony optimization, (2) mathematical pre-processing selection by a genetic algorithm, and (3) sample selection through a distance-based procedure. Outlier detection has also been included as part of the model optimization. All the above procedures have been combined into a single algorithm, whose aim is to find the best PLS calibration model within a Monte Carlo-type philosophy. Simulated and experimental examples are employed to illustrate the success of the proposed approach. [Display omitted] •A new strategy for partial least-squares optimization is presented.•Variables, pre-processing, samples and outliers are selected.•Calibrations of near infrared spectra are improved.