On sparsity, power-law and clustering properties of graphex processes Permalink

Published in Published in Advances of Applied Probability, 2023

Fran├žois Caron, Francesca Panero and Judith Rousseau

This paper investigates properties of the class of graphs based on exchangeable point processes. We talk a lot about sparsity levels, degree distributions (power-law), positive clustering coefficients and central limit theorems. And we show how these properties hold for many many models. See paper here Read more

Achieving Fairness with a Simple Ridge Penalty Permalink

Published in Statistics and Computing, 2022

Marco Scutari, Francesca Panero, Manuel Proissl

In this paper we present a general framework for estimating regression models subject to a user-defined level of fairness. We do that using ridge regression, which is flexible, has already tons of literature and in general simplifies a lot the estimation. See paper here Read more

Optimal disclosure risk assessment Permalink

Published in Annals of Statistics, 2021

Federico Camerlenghi, Stefano Favaro, Zacharie Naulet, Francesca Panero

We propose a nonparametric estimator of disclosure risk and prove its minimax optimality. It is nice because we close an open problem in the literature (20+ years!) and the optimality proof is kind of crazy (credits to Zachary). See paper here Read more