The average likelihood ratio for large-scale multiple testing and detecting sparse mixtures

DOI10.1214/12-IMSCOLL923zbMATH Open1356.62095arXiv1111.0328MaRDI QIDQ5499697FDOQ5499697

Authors: Guenther Walther 

Publication date: 30 July 2015

Published in: Institute of Mathematical Statistics Collections (Search for Journal in Brave)

Abstract: Large-scale multiple testing problems require the simultaneous assessment of many p-values. This paper compares several methods to assess the evidence in multiple binomial counts of p-values: the maximum of the binomial counts after standardization (the `higher-criticism statistic'), the maximum of the binomial counts after a log-likelihood ratio transformation (the `Berk-Jones statistic'), and a newly introduced average of the binomial counts after a likelihood ratio transformation. Simulations show that the higher criticism statistic has a superior performance to the Berk-Jones statistic in the case of very sparse alternatives (sparsity coefficient $$), while the situation is reversed for $$. The average likelihood ratio is found to combine the favorable performance of higher criticism in the very sparse case with that of the Berk-Jones statistic in the less sparse case and thus appears to dominate both statistics. Some asymptotic optimality theory is considered but found to set in too slowly to illuminate the above findings, at least for sample sizes up to one million. In contrast, asymptotic approximations to the critical values of the Berk-Jones statistic that have been developed by Wellner and Koltchinskii (2003) and Jager and Wellner (2007) are found to give surprisingly accurate approximations even for quite small sample sizes.

Full work available at URL: https://arxiv.org/abs/1111.0328

Recommendations

• Higher criticism: \(p\)-values and criticism
• Higher criticism for detecting sparse heterogeneous mixtures.
• The log-likelihood ratio for sparse multinomial mixtures
• Beyond HC: more sensitive tests for rare/weak alternatives
• Extended Likelihood Approach to Large-Scale Multiple Testing

zbMATH Keywords

higher criticismaverage likelihood ratiosparse mixtureBerk-Jones statisticasymptotic optimality theorylog-likelihood ratio transformation

Mathematics Subject Classification ID

Nonparametric hypothesis testing (62G10) Asymptotic distribution theory in statistics (62E20) Approximations to statistical distributions (nonasymptotic) (62E17) Paired and multiple comparisons; multiple testing (62J15)

Cited In (9)

• Beyond HC: more sensitive tests for rare/weak alternatives
• Overparameterized maximum likelihood tests for detection of sparse vectors
• Higher criticism: \(p\)-values and criticism
• On the exact Berk-Jones statistics and their \(p\)-value calculation
• Large-scale inference with block structure
• Genetic variant set-based tests using the generalized Berk-Jones statistic with application to a genome-wide association study of breast cancer
• Large sample bias of maximum likeihood estimates from a series of sparse ixj tables
• Higher criticism for large-scale inference, especially for rare and weak effects
• Optimal detection of multi-sample aligned sparse signals

This page was built for publication: The average likelihood ratio for large-scale multiple testing and detecting sparse mixtures

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5499697)