From Distance Correlation to Multiscale Graph Correlation
From MaRDI portal
(Redirected from Publication:3304853)
Abstract: Understanding and developing a correlation measure that can detect general dependencies is not only imperative to statistics and machine learning, but also crucial to general scientific discovery in the big data age. In this paper, we establish a new framework that generalizes distance correlation --- a correlation measure that was recently proposed and shown to be universally consistent for dependence testing against all joint distributions of finite moments --- to the Multiscale Graph Correlation (MGC). By utilizing the characteristic functions and incorporating the nearest neighbor machinery, we formalize the population version of local distance correlations, define the optimal scale in a given dependency, and name the optimal local correlation as MGC. The new theoretical framework motivates a theoretically sound Sample MGC and allows a number of desirable properties to be proved, including the universal consistency, convergence and almost unbiasedness of the sample version. The advantages of MGC are illustrated via a comprehensive set of simulations with linear, nonlinear, univariate, multivariate, and noisy dependencies, where it loses almost no power in monotone dependencies while achieving better performance in general dependencies, compared to distance correlation and other popular methods.
Recommendations
Cites work
- A consistent multivariate test of association based on ranks of distances
- Asymptotic local efficiency of Cramér\,-\,von Mises tests for multivariate independence
- Consistent distribution-free K-sample and independence tests for univariate random variables
- Consistent nonparametric tests of independence
- Detecting novel associations in large data sets
- Distance covariance in metric spaces
- Energy statistics: a class of statistics based on distances
- Entropy-based independence test
- Equivalence of distance-based and RKHS-based statistics in hypothesis testing
- Feature screening via distance correlation learning
- Kernel methods for measuring independence
- Laplacian Eigenmaps for Dimensionality Reduction and Data Representation
- Local efficiency of a Cramér\,-\,von Mises test of independence
- Measuring and testing dependence by correlation of distances
- Measuring nonlinear dependence in time-series, a distance correlation approach
- Partial distance correlation with methods for dissimilarities
- Permutation, parametric and bootstrap tests of hypotheses.
- Rejoinder: ``Brownian distance covariance
- Tests of independence among continuous random vectors based on Cramér-von Mises functionals of the empirical copula process
- The distance correlation \(t\)-test of independence in high dimension
Cited in
(24)- The exact equivalence of distance and kernel methods in hypothesis testing
- On the interpretation of inflated correlation path weights in concentration graphs
- Fourier-type tests of mutual independence between functional time series
- Graph pseudometrics from a topological point of view
- Testing independence of functional variables by angle covariance
- Distance-covariance-based tests for heteroscedasticity in nonlinear regressions
- A kernel-based measure for conditional mean dependence
- Testing serial independence with functional data
- Independent block identification in multivariate time series
- Power Analysis of Projection-Pursuit Independence Tests
- The Chi-Square Test of Distance Correlation
- A Semiparametric Kernel Independence Test With Application to Mutational Signatures
- Rank-based indices for testing independence between two high-dimensional vectors
- Extremal properties of evolving networks: local dependence and heavy tails
- Feature screening for ultra-high-dimensional data via multiscale graph correlation
- Universally consistent \(\mathrm{K}\)-sample tests via dependence measures
- Dissecting Gene Expression Heterogeneity: Generalized Pearson Correlation Squares and the K -Lines Clustering Algorithm
- An adaptive test based on Kendall's tau for independence in high dimensions
- High-Dimensional Block Diagonal Covariance Structure Detection Using Singular Vectors
- Convexity and measures of statistical association
- Active learning for high-dimensional reliability using matrix-operation anisotropic RBF model embedded with multiscale graph correlation
- Detect complete dependence via trace correlation in the presence of matrix-valued random objects
- AUK-based test for mutual independence and an index of mutual dependence
- A novel multiscale approach to independence testing
This page was built for publication: From Distance Correlation to Multiscale Graph Correlation
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3304853)
