Abstract:We propose a framework for analyzing and comparing distributions without imposing any parametric assumptions via empirical likelihood methods. Our framework is used to study two fundamental statistical test problems: the two-sample test and the goodness-of-fit test. For the two-sample test, we need to determine whether two groups of samples are from different distributions; for the goodness-of-fit test, we examine how likely it is that a set of samples is generated from a known target distribution. Specifically, we propose empirical likelihood ratio (ELR) statistics for the two-sample test and the goodness-of-fit test, both of which are of linear time complexity and show higher power (i.e., the probability of correctly rejecting the null hypothesis) than the existing linear statistics for high-dimensional data. We prove the nonparametric Wilks' theorems for the ELR statistics, which illustrate that the limiting distributions of the proposed ELR statistics are chi-square distributions. With these limiting distributions, we can avoid bootstraps or simulations to determine the threshold for rejecting the null hypothesis, which makes the ELR statistics more efficient than the recently proposed linear statistic, finite set Stein discrepancy (FSSD). We also prove the consistency of the ELR statistics, which guarantees that the test power goes to 1 as the number of samples goes to infinity. In addition, we experimentally demonstrate and theoretically analyze that FSSD has poor performance or even fails to test for high-dimensional data. Finally, we conduct a series of experiments to evaluate the performance of our ELR statistics as compared to state-of-the-art linear statistics.

Log-Rank-Type Tests for Equality of Distributions in High-Dimensional Spaces.

The Asymptotic Distribution and Berry–Esseen Bound of a New Test for Independence in High Dimension with an Application to Stochastic Optimization

A Rank-Based High-Dimensional Test for Equality of Mean Vectors

Nonparametric Two-Sample Tests for Equality of Distributions Using Projections

Non-parametric Tests for the Tail Equivalence Via Empirical Likelihood

A Robust and Nonparametric Two-Sample Test in High Dimensions

Identity Tests for High Dimensional Data Using Rmt

Bayesian Optimal Two-sample Tests in High-dimension

Two-sample Test in High Dimensions Through Random Selection.

RANK-BASED INDICES FOR TESTING INDEPENDENCE BETWEEN TWO HIGH-DIMENSIONAL VECTORS

Multivariate Rank-Based Distribution-Free Nonparametric Testing Using Measure Transportation

Two-Sample Test for Sparse High Dimensional Multinomial Distributions

Linear Kernel Tests Via Empirical Likelihood for High-Dimensional Data

Testing high-dimensional multinomials with applications to text analysis

Two-Sample Smooth Tests for the Equality of Distributions

Joint Test for Homogeneity of High-Dimensional Means and Covariance Matrices Using Maximum-Type Statistics

A Hyperbolic Divergence Based Nonparametric Test for Two‐sample Multivariate Distributions

Exact and efficient multivariate two-sample tests through generalized linear rank statistics

Testing distributional equality for functional random variables

An RIHT Statistic for Testing the Equality of Several High-Dimensional Mean Vectors under Homoskedasticity

On testing the equality of high dimensional mean vectors with unequal covariance matrices