# Kernel-based Tests for Joint Independence

Source: [Kernel-based Tests for Joint Independence](<https://www.sota2.com/research/paper/68fa53501e209131ce982727>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T12:23:09.447Z

Authors: Niklas Pfister, Peter B\\"uhlmann, Bernhard Sch\\"olkopf, Jonas Peters

Published: 2016-03-01T14:34:53.000Z

## Abstract

We investigate the problem of testing whether $d$ random variables, which may or may not be continuous, are jointly (or mutually) independent. Our method builds on ideas of the two variable Hilbert-Schmidt independence criterion (HSIC) but allows for an arbitrary number of variables. We embed the $d$-dimensional joint distribution and the product of the marginals into a reproducing kernel Hilbert space and define the $d$-variable Hilbert-Schmidt independence criterion (dHSIC) as the squared distance between the embeddings. In the population case, the value of dHSIC is zero if and only if the $d$ variables are jointly independent, as long as the kernel is characteristic. Based on an empirical estimate of dHSIC, we define three different non-parametric hypothesis tests: a permutation test, a bootstrap test and a test based on a Gamma approximation. We prove that the permutation test achieves the significance level and that the bootstrap test achieves pointwise asymptotic significance level as well as pointwise asymptotic consistency (i.e., it is able to detect any type of fixed dependence in the large sample limit). The Gamma approximation does not come with these guarantees; however, it is computationally very fast and for small $d$, it performs well in practice. Finally, we apply the test to a problem in causal discovery.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Independence testing on IndependentClouds (test)](<https://www.sota2.com/research/sota/independence-testing-on-independentclouds-test.md>): Independence testing; IndependentClouds (test); Rejection Rate: 4.

- [Independence testing on Diamond distribution (test)](<https://www.sota2.com/research/sota/independence-testing-on-diamond-distribution-test.md>): Independence testing; Diamond distribution (test); Rejection Rate: 100.

- [Independence testing on Parabola distribution (test)](<https://www.sota2.com/research/sota/independence-testing-on-parabola-distribution-test.md>): Independence testing; Parabola distribution (test); Rejection Rate: 1.

- [Independence testing on TwoParabola distribution (test)](<https://www.sota2.com/research/sota/independence-testing-on-twoparabola-distribution-test.md>): Independence testing; TwoParabola distribution (test); Rejection Rate: 100.

- [Independence testing on Circle distribution (test)](<https://www.sota2.com/research/sota/independence-testing-on-circle-distribution-test.md>): Independence testing; Circle distribution (test); Rejection Rate: 100.

- [Independence testing on Variance distribution (test)](<https://www.sota2.com/research/sota/independence-testing-on-variance-distribution-test.md>): Independence testing; Variance distribution (test); Rejection Rate: 100.

- [Independence testing on Log distribution (test)](<https://www.sota2.com/research/sota/independence-testing-on-log-distribution-test.md>): Independence testing; Log distribution (test); Rejection Rate: 100.

- [Independence testing on W distribution (test)](<https://www.sota2.com/research/sota/independence-testing-on-w-distribution-test.md>): Independence testing; W distribution (test); Rejection Rate: 1.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/1603.00285>)

[PDF](<https://arxiv.org/pdf/1603.00285.pdf>)
