Variable selection methods, such as heuristic searches
including genetic algorithms (GA), simulated annealing algorithms
and particle swarm optimization (PSO), could be utilized to improve
the predictive and computational efficiency of the calibration
models by selecting optimal wavelengths. In this study, the
PLS-Genetic algorithm toolbox (Leardi and Lupiáñez, 1998) was
used to select optimal wavelengths; all the values in this toolbox
were default values, except the number of runs was set to 500.