MODEL
The model is semi-automatically selected based on the imported data formats. This means that if three-way data (X) is imported and preserved during pre-processing, then PARAFAC, PARAFAC2 or N-PLS can be selected – these models will then be highlighted in the “model” menu function. Otherwise if the feature extraction or the exclusion of samples, times or sensors reduces the dimension to a two-way matrix then a PCA, PLS1 or PLS2 model can be calculated. The type of PLS model depends of the size of the Y-matrix imported.
If a PARAFAC2 model is calculated the shifted profiles must be located within the second mode (= time mode) and across the last mode (sample or sensor mode) according to the convention of the model given in the PARAFAC2 toolbox (http://www.models.kvl.dk/source/).
Set parameters
This menu allows you to set how many components that should be calculated in the chemometric model. It also allows you to decide whether to use cross-validation or not.
Include/exclude
Y-variables
Allows you to exclude one or more of your Y-variables. This may be advisable if you want to investigate the effect of a single variable (PLS model) or one variable disturbs the interpretation of the correlation between the other Y-variables and the X-matrix.
PCA
Works for two-way data. More information can be found in:
- Wold, S.; Esbensen, K. & Geladi, P. (1987). Principal Component Analysis. Chemometrics and Intelligent Laboratory Systems, vol. 2: 37-52.
PARAFAC
Works for three-way data. More information can be found in:
- Bro, R. (1997). PARAFAC – Tutorial and Applications. Chemometrics and Intelligent Laboratory Systems, vol. 38: 149-171.
PARAFAC2
Works for three-way data. More information can be found in:
-
Bro, R.; Kiers,
- Kiers, H.A.L.; Berge, ten J.M.F. & Bro, R. (1999). PARAFAC2 - Part I. A direct fitting algorithm for the PARAFAC2 model. Journal of Chemometrics, vol. 13: 275-294.
PLS
Works for two-way data with both X and Y variables. More information can be found in:
- Wold, S.; Sjostrom, M. & Eriksson, L. (2001). PLS-regression: a basic tool of chemometrics. Chemometrics and Intelligent Laboratory Systems, vol. 58, no. 2: 109-130.
N-PLS
Works for three-way data with both X and Y variables. More information can be found in:
-
Bro, R. (1996). Multi-way calibration. Multi-linear PLS. Journal of Chemometrics, vol. 10(1):
47-62.