arXiv subjects

Romain Périer

Publications and source records attributed to Romain Périer.

3 recordsLinked to original sources

arXiv2026

On strongly coherent confidence envelopes for the false discoveries

Coherence is a long term notion in the literature of multiple testing. From FWER control multiple testing theory to confidence envelopes on the number of false positives, coherence is a desirable property for multiple testing procedures. However, we show in this article that coherence is not a sufficient property to ensure that a multiple testing procedure is nonimprovable via an interpolation mechanism. We introduce a new no- tion, the notion of strong coherence, which is a stronger property than coherence and ensures that a multiple testing procedure is nonimprovable via the interpolation mech- anisms considered in the literature. Finally, we provide a uniform improvement of the confidence envelope from Durand et al. (2020) that is computable in polynomial time and demonstrate the improvement on simulated data.

math.ST↗

arXiv2026

Confidence envelopes for the false discoveries with heterogeneous data

In the context of selective inference, confidence envelopes for the false discoveries allow the user to select any subset of null hypotheses while having a statistical guarantee on the number of false discoveries in the selected set. Many constructions of such envelopes have been proposed recently, using local test families (Genovese and Wasserman, 2006; Goeman and Solari, 2011), paths (Katsevich and Ramdas, 2020) or interpolation (Blanchard et al., 2020a). All those methods have in common that they have been well-studied for the homogeneous case where all p-values under the null have a uniform distribution over [0, 1]. However, in many applications the data are heterogeneous and discrete, hence the p-values have heterogeneous, discrete distributions, and the previous constructions may incur a loss of power, in the sense that they over-estimate the number of false discoveries. In this paper, we bridge the previous constructions under the homogeneous case with new tools. We also apply these tools to propose several confidence envelopes based on tools tailored for heterogeneous data, like the Bretagnolle inequality, or a new variant of the Simes inequality. We compare these new envelopes to their homogeneous counterparts on simulated data.

math.ST↗

arXiv2024

FDR control and FDP bounds for conformal link prediction

In Marandon (2023), the author introduces a procedure to detect true edges from a partially observed graph using a conformal prediction fashion: first computing scores from a trained function, deriving conformal p-values from them and finally applying a multiple testing procedure. In this paper, we prove that the resulting procedure indeed controls the FDR, and we also derive uniform FDP bounds, thanks to an exchangeability argument and the previous work of Marandon et al. (2022).

math.ST↗