A0738
Title: Empirical Bayes large-scale multiple testing for high-dimensional binary outcome data
Authors: Yu-Chien Bo Ning - Harvard T.H. Chan Public Health (United States) [presenting]
Abstract: The purpose is to discuss a recent result on the multiple testing problem for sparse high-dimensional data with binary outcomes. A novel empirical Bayes multiple testing procedures, based on a spike-and-slab posterior is introduced, and their performance in controlling the false discovery rate (FDR) is evaluate. A surprising finding is that the procedure using the default conjugate prior (namely, the -value procedure) can be overly conservative in estimating the FDR. To address this, two new procedures are introduced that both provide accurate FDR control. Sharp frequentist theoretical results are established for these procedures, and numerical experiments are conducted to validate the theory in finite samples. To the best of knowledge, this is the first work that obtains uniform FDR control results in multiple testing for high-dimensional data with binary outcomes under the sparsity assumption.
