A comparison of multivariate calibration techniques applied to experimental NIR data sets: Part III: Robustness against instrumental perturbation conditions

This work is part of a more general research aiming at comparing the performance of multivariate calibration methods. In the first and second parts of the study, the performances of multivariate calibration methods were compared in situations of interpolation and extrapolation, respectively. This third part of the study deals with robustness of calibration methods in the case where spectra corresponding to new samples of which the y value has to be predicted are affected by instrumental perturbations not accounted for in the calibration set. These types of perturbations can happen due to instrument ageing, replacement of one or several parts of the spectrometer (e.g. the detector), use of a new instrument, or modifications in the measurement conditions, like the displacement of the instrument to a different location. Although no general rules could be extracted from the results, the variety of data sets and methods tested allowed some guidelines for multivariate calibration in this unfavourable case of instrumental perturbation to be given. Models based on Neural Networks (NN) applied to Fourier Coefficient proved particularly robust in some cases, and failed badly in some others. The most stable methods were principal component regression (PCR, with component selection) and partial least squared regression (with complexity optimisation performed by randomisation test).