Prof. H. Leeb (University of Vienna, Österreich)
Wednesday, May 19, 2021 - 10:00
Online Event
Der Vortrag findet bei Zoom statt: https://zoom.us/j/159082384, --- ---
Forschungsseminar Mathematische Statistik
We show that two popular selective inference procedures, namely data carving (Fithian et al., 2017) and selection with a randomized response (Tian et al., 2018b), when combined with the polyhedral method (Lee et al., 2016), result in confidence intervals whose length is bounded. This contrasts results for confidence intervals based on the polyhedral method alone, whose expected length is typically infinite (Kivaranovic and Leeb, 2020). Moreover, we show that these two procedures always dominate corresponding sample-splitting methods in terms of interval length.
submitted by chschnei (christine.schneider@wias-berlin.de, 030 20372574)