Kernel-based tests for joint independence

DOI10.1111/RSSB.12235MaRDI QIDQ4603812zbMATH OpenOpenAlexWikidataFDO

Authors Niklas Pfister, Bernhard Schölkopf, Jonas Peters, Peter Bühlmann

Publication date 19 February 2018

Published in Journal of the Royal Statistical Society Series B: Statistical Methodology (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1603.00285, https://curis.ku.dk/portal/da/publications/kernelbased-tests-for-joint-independence(d25bedca-a0df-4d9e-94d5-a7e2c6aa00ad).html

zbMATH Keywords

causal inferenceindependence testkernel methodsV-statistics

Mathematics Subject Classification ID

Nonparametric hypothesis testing (62G10) Estimation in multivariate analysis (62H12)

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.

Recommendations

• Consistent nonparametric tests of independence
• Nonparametric Independence Tests: Space Partitioning and Kernel Approaches
• A multivariate nonparametric test of independence
• Algorithmic Learning Theory
• Multivariate tests of independence based on a new class of measures of independence in reproducing kernel Hilbert space

Cites work

• A Bootstrap Revival of Some Nonparametric Distance Tests
• A Consistent Test for Bivariate Dependence
• A consistent test of independence based on a sign covariance related to Kendall's tau
• A Hilbert Space Embedding for Distributions
• A kernel two-sample test
• A multivariate nonparametric test of independence
• A multivariate version of Hoeffding's phi-square
• Adaptive estimation of the copula correlation matrix for semiparametric elliptical copulas
• An introduction to sparse stochastic processes
• Approximation Theorems of Mathematical Statistics
• Bootstrap and randomization tests of some nonparametric hypotheses
• Brownian distance covariance
• CAM: causal additive models, high-dimensional order search and penalized regression
• Distinguishing cause from effect using observational data: methods and benchmarks
• Efficient independent component analysis
• High-dimensional semiparametric Gaussian copula graphical models
• scientific article; zbMATH DE number 6378135 (Why is no real title available?)
• Large-Sample Theory for the Bergsma-Dassios Sign Covariance
• Partial distance correlation with methods for dissimilarities
• Regularized rank-based estimation of high-dimensional nonparanormal graphical models
• Stochastic estimation and testing
• Structural intervention distance for evaluating causal graphs

Cited in

(93)

• A fast algorithm for computing distance correlation
• A new coefficient of correlation
• Optimal rates for independence testing via U-statistic permutation tests
• Generalizing distance covariance to measure and test multivariate mutual dependence via complete and incomplete V-statistics
• Large-scale kernel methods for independence testing
• Independence test for large sparse contingency tables based on distance correlation
• High-dimensional consistent independence testing with maxima of rank correlations
• Some tests of independence based on maximum mean discrepancy and ranks of nearest neighbors
• The exact equivalence of distance and kernel methods in hypothesis testing
• Statistical dependence: beyond Pearson's
• Some new copula based distribution-free tests of independence among several random variables
• A regression perspective on generalized distance covariance and the Hilbert-Schmidt independence criterion
• On the asymptotic null distribution of the symmetrized Chatterjee's correlation coefficient
• Adaptive test of independence based on HSIC measures
• Multivariate ranks and quantiles using optimal transport: consistency, rates and nonparametric testing
• Grouped feature importance and combined features effect plot
• On some consistent tests of mutual independence among several random vectors of arbitrary dimensions
• Testing independence of functional variables by angle covariance
• A kernel-based measure for conditional mean dependence
• Tests of zero correlation using modified RV coefficient for high-dimensional vectors
• Distance multivariance: new dependence measures for random vectors
• WIKS: a general Bayesian nonparametric index for quantifying differences between two populations
• Multiple testing of conditional independence hypotheses using information-theoretic approach
• Multivariate tests of independence based on a new class of measures of independence in reproducing kernel Hilbert space
• Kernel methods for independence measurement with coefficient constraints
• Consistent nonparametric tests of independence
• Nonparametric Independence Tests: Space Partitioning and Kernel Approaches
• Some blum-kiefer-rosenblatt type tests for the joint independence of variables
• Characteristic and universal tensor product kernels
• Tests of mutual or serial independence of random vectors with applications
• New HSIC-based tests for independence between two stationary multivariate time series
• Tests of mutual independence among several random vectors using univariate and multivariate ranks of nearest neighbours
• Comparing two populations using Bayesian Fourier series density estimation
• Copula versions of distance multivariance and dHSIC via the distributional transform -- a general approach to construct invariant dependence measures
• Expected conditional characteristic function-based measures for testing independence
• BET on independence
• Distance metrics for measuring joint dependence with application to causal inference
• Composite coefficient of determination and its application in ultrahigh dimensional variable screening
• Invariant causal prediction for sequential data
• Toward Causal Inference for Spatio-Temporal Data: Conflict and Forest Loss in Colombia
• The Hellinger Correlation
• Generalization of the HSIC and distance covariance using PDI kernels
• A distribution-free test of independence based on a modified mean variance index
• The Binary Expansion Randomized Ensemble Test
• An Updated Literature Review of Distance Correlation and Its Applications to Time Series
• Some correlation tests for vectors of large dimension
• Test of bivariate independence based on angular probability integral transform with emphasis on circular-circular and circular-linear data
• Distributed testing on mutual independence of massive multivariate data
• Test of conditional independence in factor models via Hilbert-Schmidt independence criterion
• From statistical to causal learning
• Testing independence of functional variables by an Hilbert-Schmidt independence criterion estimator
• HSIC regularized LTSA
• Some copula-based tests of independence among several random variables having arbitrary probability distributions
• Consistency of permutation tests of independence using distance covariance, HSIC and dHSIC
• Test of independence for Hilbertian random variables
• Statistical Inferences for Complex Dependence of Multimodal Imaging Data
• Inferring the finest pattern of mutual independence from data
• A kernel- and optimal transport- based test of independence between covariates and right-censored lifetimes
• A survey of some recent developments in measures of association
• A conditional distribution function-based measure for independence and K-sample tests in multivariate data
• Power boosting: fusion of multiple test statistics via resampling
• Distributed testing on mutual independence between components of high-dimensional massive data
• Kernel-based marginal testing for covariate effects in high-dimensional settings
• Testing for independence in high dimensions based on characteristic covariance
• Instrumental variable estimation of distributional causal effects
• A kernel-based approach for testing mutual independence of several functional variables
• Rank test of independence based on adaptive Hellinger distance
• Sample splitting and assessing goodness-of-fit of time series
• Enhanced HSIC for independence test via projection integration
• Measuring dependence between functional data via projection Hilbert-Schmidt covariance
• A kernel-based test for the first-order separability of spatio-temporal point processes
• A distance covariance test of independence in high dimension, low sample size contexts
• An adaptive test of the independence of high-dimensional data based on Kendall rank correlation coefficient
• On Exact Feature Screening in Ultrahigh-Dimensional Binary Classification
• Using Rejection Sampling Probability of Acceptance as a Measure of Independence
• Kernel Angle Dependence Measures in Metric Spaces
• Studentized tests of independence: random-lifter approach
• Kernel copula density estimation of Hellinger correlation
• A kernel independence test using projection-based measure in high-dimension
• Specification procedures for multivariate stable-Paretian laws for independent and for conditionally heteroskedastic data
• A nonparametric distribution-free test of independence among continuous random vectors based on L₁-norm
• Kernel-based method for joint independence of functional variables
• On relationships between Chatterjee’s and Spearman’s correlation coefficients
• AUK-based test for mutual independence and an index of mutual dependence
• Structural restrictions in local causal discovery: identifying direct causes of a target variable
• A nonparametric test of multivariate independence using graphs
• Test of independence using generalized distance correlation
• Distribution-free robust independence test for high-dimensional data via semi-Grothendieck's covariance
• Identifiability of causal graphs under non-additive conditionally parametric causal models
• DiPMInd: distance profile based mutual independence testing for random objects
• Robust independence test of functional variables via Gupta angle covariance
• A novel multiscale approach to independence testing
• A Graph-Based Framework for Nonparametric Tests of Multivariate Independence

This page was built for publication: Kernel-based tests for joint independence

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4603812)