Abstract:Consider a situation of analyzing high-dimensional count data containing an excess amount of zeroes and small non-zero counts. Under the assumption that the observations are modeled by a Poisson distribution, we are interested in simultaneous testing of the means of those observations. In this work, we study some optimal properties (in terms of Bayes risk) of multiple-testing rules induced by both two-groups and a general class of one-group shrinkage priors. The class of one-group priors was proposed by Polson and Scott (2010) and studied by Ghosh et al. (2016) in the context of the normal means model and by Tang et al. (2018) for the linear regression model. Here, first, we model each mean by a two-groups prior, and under the assumption of $0-1$ loss function, we obtain an expression for the optimal Bayes risk under some assumption similar to Bogdan et al. (2011). Next, same as Ghosh et al. (2016), we assume a two-groups mixture model for the data and model each mean parameter by the broad class of one-group priors to study the Bayes risk induced by the chosen class of priors. We have been able to show that, when the underlying sparsity pattern is known, under some proposed assumptions, the Bayes risk corresponding to the broad class of priors attains the optimal Bayes risk exactly. In other words, it ensures the decision rule is Asymptotically Bayes optimal under sparsity (ABOS). When this sparsity is unknown, motivated by Yano et al. (2021), we use an empirical Bayes estimate of the global shrinkage parameter. In this case, also, we show that the modified decision rule is ABOS, too. In this way, as an alternative solution for two-groups prior, we propose a broad class of global-local priors having similar optimal properties in terms of Bayes risk to quasi-sparse count data.

The Minimax Risk in Testing Uniformity of Poisson Data under Missing Ball Alternatives within a Hypercube

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

Locally sharp goodness-of-fit testing in sup norm for high-dimensional counts

Testing the homogeneity of risk differences with sparse count data

Robust Hypothesis Testing for Asymmetric Nominal Densities under a Relative Entropy Tolerance

A NEW PERSPECTIVE ON ROBUST M-ESTIMATION: FINITE SAMPLE THEORY AND APPLICATIONS TO DEPENDENCE-ADJUSTED MULTIPLE TESTING

On uniformly consistent tests

Two-Sample Test for Sparse High Dimensional Multinomial Distributions

Testing Poisson Binomial Distributions

Asymptotic Bayes Optiamlity for Sparse Count Data

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

Robust mean change point testing in high-dimensional data with heavy tails

Nearly Minimax Optimal Wasserstein Conditional Independence Testing

Asymptotic bayes optimality under sparsity for equicorrelated multivariate normal test statistics

Minimax Sequential Testing for Poisson Processes

A Robust and Nonparametric Two-Sample Test in High Dimensions

Approximative Tests for the Equality of Two Cumulative Incidence Functions of a Competing Risk

Testing the Isotropic Cauchy Hypothesis

Universal Hypothesis Testing with Kernels: Asymptotically Optimal Tests for Goodness of Fit

Asymptotics for Outlier Hypothesis Testing

On High dimensional Poisson models with measurement error: hypothesis testing for nonlinear nonconvex optimization