Controlling the False Discovery Rate via Competition: is the +1 needed?
arXiv:2204.13248 · doi:10.1016/j.spl.2023.109819
Abstract
Barber and Candès (2015) control of the FDR in feature selection relies on estimating the FDR by the number of knockoff wins +1 divided by the number of original wins. We study the necessity of the +1 in general settings.
10 pages + 9 supplementary pages