ACOGASS: variable, sample and pre-processing selection

Graphical user interface for the simultaneous selection of variables, samples and pre-processing methods in multivariate calibration.

ACOGASS is a MATLAB graphical user interface for the simultaneous selection of calibration/validation samples, spectral (or other types of) variables and mathematical pre-processing methods in multivariate calibration with partial leas-squares (PLS-1). The interface integrates different algorithms for the selections: (1) ant colony optimization for variables, (2) genetic algorithms for pre-processing, and (3) Kennard-Stone and SPXY for samples.