[Statlist] Research Seminar in Statistics | *FRIDAY, OCTOBER 17, 2025* | GSEM, University of Geneva

[gsem-support-instituts]

Dear all,

We are pleased to invite you to our next Research Seminar, organized by Professor Davide La Vecchia on behalf of the Research Institute for Statistics and Information Science
< https://www.unige.ch/gsem/en/research/institutes/risis/team/ >.

FRIDAY, OCTOBER 17, 2025, at 11:15 am, Uni Mail M 5220.

Robust High-Dimensional Linear Regression: Convex and Non-Convex Estimators for Adversarial and Heavy-Tailed Settings

Mohamed NDAOUD, ESSEC Business School, France
< https://faculty.essec.edu/cv/ndaoud-mohamed/ >

Based on the papers:

*Robust and Tuning-Free Sparse Linear Regression via Square-Root Slope.
Minsker, S., Ndaoud, M. & Wang, L.

*Outlier-Bias Removal with Alpha Divergence: A Robust Non-Convex Estimator for Linear Regression.
Hammouda I., Ndaoud, M. & Seghouane Abd K.

ABSTRACT:
We study high-dimensional linear regression under adversarial contamination and heavy-tailed noise. We establish minimax lower bounds that capture the effect of both corruption and noise tails, and we propose two complementary estimators. The first is a convex, tuning-free variant of the Square-Root Slope that is adaptive, computationally efficient, and minimax optimal up to logarithmic factors. The second is a non-convex M-estimator based on alpha divergence that removes outlier-induced bias and achieves minimax-optimal deviation bounds, improved robustness to large outliers, and asymptotic normality. Together, these approaches provide practical and theoretically grounded methods for robust high-dimensional regression in adversarial and heavy-tailed settings.

> View the Research Seminar agenda: < https://www.unige.ch/gsem/en/research/seminars/risis/ >

Regards,
Birgit

Birgit Müller-Marreros
Assistant to the Research Institutes
gsem-support-instituts at unige.ch

GSEM Support for Research Institutes
Office Uni Mail 3266 | +41 22 379 88 13
40 bd du Pont d'Arve | CH - 1211 Geneva 4