Kernel Constrained Covariance for Dependence Measurement
Abstract
We discuss reproducing kernel Hilbert space (RKHS)-based measures of statistical dependence, with emphasis on constrained covariance (COCO), a novel criterion to test dependence of random variables. We show that COCO is a test for independence if and only if the associated RKHSs are universal. That said, no independence test exists that can distinguish dependent and independent random variables in all circumstances. Dependent random variables can result in a COCO which is arbitrarily close to zero when the source densities are highly non-smooth. All current kernel-based independence tests share this behaviour. We demonstrate exponential convergence between the population and empirical COCO. Finally, we use COCO as a measure of joint neural activity between voxels in MRI recordings of the macaque monkey, and compare the results to the mutual information and the correlation. We also show the effect of removing breathing artefacts from the MRI recording.
Cite
Text
Gretton et al. "Kernel Constrained Covariance for Dependence Measurement." Proceedings of the Tenth International Workshop on Artificial Intelligence and Statistics, 2005.Markdown
[Gretton et al. "Kernel Constrained Covariance for Dependence Measurement." Proceedings of the Tenth International Workshop on Artificial Intelligence and Statistics, 2005.](https://mlanthology.org/aistats/2005/gretton2005aistats-kernel/)BibTeX
@inproceedings{gretton2005aistats-kernel,
title = {{Kernel Constrained Covariance for Dependence Measurement}},
author = {Gretton, Arthur and Smola, Alexander and Bousquet, Olivier and Herbrich, Ralf and Belitski, Andrei and Augath, Mark and Murayama, Yusuke and Pauls, Jon and Schölkopf, Bernhard and Logothetis, Nikos},
booktitle = {Proceedings of the Tenth International Workshop on Artificial Intelligence and Statistics},
year = {2005},
pages = {112-119},
volume = {R5},
url = {https://mlanthology.org/aistats/2005/gretton2005aistats-kernel/}
}