Large-scale kernel methods for independence testing

DOI10.1007/S11222-016-9721-7zbMATH Open1384.62154arXiv1606.07892OpenAlexW2463154922MaRDI QIDQ1702289FDOQ1702289

Authors: Qinyi Zhang, S. Filippi, Arthur Gretton, Dino Sejdinovic 

Publication date: 28 February 2018

Published in: Statistics and Computing (Search for Journal in Brave)

Abstract: Representations of probability measures in reproducing kernel Hilbert spaces provide a flexible framework for fully nonparametric hypothesis tests of independence, which can capture any type of departure from independence, including nonlinear associations and multivariate interactions. However, these approaches come with an at least quadratic computational cost in the number of observations, which can be prohibitive in many applications. Arguably, it is exactly in such large-scale datasets that capturing any type of dependence is of interest, so striking a favourable tradeoff between computational efficiency and test performance for kernel independence tests would have a direct impact on their applicability in practice. In this contribution, we provide an extensive study of the use of large-scale kernel approximations in the context of independence testing, contrasting block-based, Nystrom and random Fourier feature approaches. Through a variety of synthetic data experiments, it is demonstrated that our novel large scale methods give comparable performance with existing methods whilst using significantly less computation time and memory.

Full work available at URL: https://arxiv.org/abs/1606.07892

Recommendations

• Kernel methods for measuring independence
• Kernel-based tests for joint independence
• Testing independence by nonparametric kernel method
• Kernel methods for independence measurement with coefficient constraints
• Testing the independence of sets of large-dimensional variables
• Testing independence among a large number of high-dimensional random vectors
• Large-scale simultaneous testing using kernel density estimation
• Testing for independence of large dimensional vectors

zbMATH Keywords

independence testing[https://portal.mardi4nfdi.de/w/index.php?title=+Special%3ASearch&search=Nystr%EF%BF%BD%EF%BF%BDm+method&go=Go Nystr��m method]Hilbert-Schmidt independence criterialarge-scale kernel methodrandom Fourier features

Mathematics Subject Classification ID

Nonparametric hypothesis testing (62G10) Learning and adaptive systems in artificial intelligence (68T05) Measures of association (correlation, canonical correlation, etc.) (62H20)

Cites Work

• FastMMD: Ensemble of Circular Discrepancy for Efficient Two-Sample Test
• Measuring and testing dependence by correlation of distances
• Equivalence of distance-based and RKHS-based statistics in hypothesis testing
• Title not available (Why is that?)
• Distance covariance in metric spaces
• Theory of Reproducing Kernels
• On the bootstrap of \(U\) and \(V\) statistics
• Two-sample test statistics for measuring discrepancies between two multivariate probability density functions using kernel-based density estimates
• 10.1162/153244303768966085
• Support Vector Machines
• A Hilbert Space Embedding for Distributions
• Title not available (Why is that?)
• Algorithmic Learning Theory
• A kernel two-sample test
• Title not available (Why is that?)
• Brownian distance covariance
• Approximation theorems of mathematical statistics
• Feature selection via dependence maximization
• Scattered Data Approximation
• Title not available (Why is that?)
• Algorithms for learning kernels based on centered alignment
• Nonlinear measures of association with kernel canonical correlation analysis and applications
• Nonlinear canonical analysis and independence tests

Cited In (17)

• A Kernel Log-Rank Test of Independence for Right-Censored Data
• Testing semiparametric model-equivalence hypotheses based on the characteristic function
• Tests of mutual or serial independence of random vectors with applications
• The Binary Expansion Randomized Ensemble Test
• Nonparametric Independence Tests: Space Partitioning and Kernel Approaches
• A kernel- and optimal transport- based test of independence between covariates and right-censored lifetimes
• New HSIC-based tests for independence between two stationary multivariate time series
• A survey of some recent developments in measures of association
• A new coefficient of correlation
• The exact equivalence of distance and kernel methods in hypothesis testing
• Statistical dependence: beyond Pearson's \(\rho\)
• A fast algorithm for computing distance correlation
• Multivariate tests of independence based on a new class of measures of independence in reproducing kernel Hilbert space
• Validation of association
• Title not available (Why is that?)
• The Chi-Square Test of Distance Correlation
• Large-scale simultaneous testing using kernel density estimation

Uses Software

• FastMMD

This page was built for publication: Large-scale kernel methods for independence testing

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1702289)