###############
###############
Files implementing the cross-validation methods described in the paper "Cross-validation techniques for determining the number of correlated components between two data sets when the number of samples is very small" by Christian Lameiro and Peter J. Schreier, Proceedings of the 50th Asilomar Conference on Signals, Systems and Computers, Pacific Grove, CA, USA, 2016
############
############
We consider the problem of estimating of the number of components that are correlated between two sets of high-dimensional data. In many applications the number of available samples is very small, in which case conventional techniques do not accurately determine the model order. Recent approaches for the sample-poor scenario are based on a combined PCA-CCA (principal component analysis-canonical correlation analysis) setup, for which they jointly determine the required model orders for the PCA and CCA steps using either information-theoretic criteria or a sequence of hypothesis tests. In this paper, we propose various alternative approaches based on cross-validation.
###########
###########
In case of questions, suggestions, problems etc. please send an email.
Christian Lameiro: [email protected]