Note that the dependencies with respect to the supervised variable are not taken into account.
A. Simulation study for the example of supervised variable selection.
Several empirical studies have investigated the extent of bias induced by performing preliminary supervised variable selection before CV.
If performed before splitting the dataset into K folds, supervised variable selection often leads to strongly downwardly biased error estimates.
As the preparation step we used supervised variable selection, which displayed the largest CVIIM s, n,K-values in the real data analyses.
In the case of the microarray data a two-way ANOVA with the factors time and dye-swap for six technical replicates was performed as described for the preselection process in preparation for supervised variable selection.
> In this illustrative analysis, through our new measure CVIIM we have confirmed the conclusion previously obtained in the literature: performing supervised variable selection before CV leads to a strong bias of the resulting error estimate.
The Minimum Redundancy Maximum Relevance (MRMR) approach to supervised variable selection represents a successful methodology for dimensionality reduction, which is suitable for high-dimensional data observed in two or more different groups.
