A1050
Title: Boosting e-BH via conditional calibration
Authors: Zhimei Ren - University of Pennsylvania (United States) [presenting]
Junu Lee - University of Pennsylvania (United States)
Abstract: The e-BH procedure is an e-value-based multiple testing procedure that provably controls the false discovery rate (FDR) under any dependence structure between the e-values. Despite this appealing theoretical FDR control guarantee, the e-BH procedure often suffers from low power in practice. A general framework is proposed that boosts the power of e-BH without sacrificing its FDR control under arbitrary dependence. This is achieved by the technique of conditional calibration, where the e-values are taken as input and are calibrated to be a set of boosted e-values that are guaranteed to be no less and are often more powerful than the original ones. The general framework is explicitly instantiated in three classes of multiple testing problems: (1) testing under parametric models, (2) conditional independence testing under the model-X setting, and (3) model-free conformalized selection. Extensive numerical experiments show that the proposed method significantly improves the power of e-BH while continuing to control the FDR.
