[Back]

D. Pérez-Guaita, J. Kuligowski, B. Lendl, B. Wood, G. Quintas:
"Assessment of discriminant models in infrared imaging using constrained repeated random sampling - Cross validation";
Analytica Chimica Acta, 1033 (2018), 156 - 164.

IR imaging is an emerging and powerful approach for studying the mol. compn. of cells and tissues. It is a nondestructive
and phenotypic technique which combines label-free mol. specific information from cells and tissues provided
by IR with spatial resoln., offering great potential in biochem. and biomedical research and routine applications. The
application of multivariate discriminant anal. using bilinear models such as Partial Least Squares-Discriminant Anal.
(PLS-DA) to IR images requires to unfold the spatial directions in a two-way matrix, resulting in a loss of spatial
information and structure. In this article, first we evidence that internal validation methods such as repeated k-fold crossvalidation
(CV) can be overly optimistic when the pixel size of the image is lower than the lateral spatial resoln.
Secondly, we propose a new approach for the unbiased internal evaluation of the model performance named
COnstrained Repeated Random Subsampling-Cross Validation (CORRS-CV). This method is based on the generation
of q training and test sub-sets using a constrained random sampling of n training pixels without replacement and it
circumvents overly optimistic effects due to oversampling, providing more accurate and robust images. The approach
can be applied in IR microscopy for the development of discriminant models to analyze underlying biochem. differences
assocd. to anatomical and histopathol. features in cells and tissues.

http://dx.doi.org/10.1016/j.aca.2018.05.019