Characterization of multi-way spectral data using factorial correspondenceregression

The principal advantage in Factorial Correspondence Analysis, where rows an d columns are processed symmetrically, is the possibility to have in the sa me factorial space observation (row) and variable (column) projections. For sequence of spec tra, the joint plot is composed of projections of wavelen gths and of spectra. In the reported study, the analyzed data set consisted in fluorescence emission spectra recorded on animal feed samples. Samples were composed of eight raw materials (4 cereals and 4 oilcakes) and 48 mixt ures of one cereal and one oilcake. For each sample, several specific excit ation wavelengths were used leading to a 3-dimensional or 3-way data set: o ne way for samples, one for emission wavelengths and one for excitation wav elengths. After reorganization of the data set. FCA was applied and the res ulting joint plot allowed finding similarities between excitation-emission wavelength couples and samples. Furthermore, the association of the Partial Least-Squares regression (PLS) with the FCA method led to the selection of some wavelength couples characteristic of the eight raw materials. The mat hematical procedure of this new regression technique, called Factorial Corr espondence Partial Least Squares regression (FCR-PLS), is developed and the model validation, which is based on a cross-validation procedure to choose independent variables entering the regression equation, is reported. All c omputations were done with Matlab(R) and programming examples are given.