4 papers
Implicit vs. explicit regularization for high-dimensional gradient descent
Thomas Stark, Lukas Steinberger
In this paper we investigate the generalization error of gradient descent (GD) applied to an -regularized OLS objective function in the linear model. Based on our analysis…
Towards multi-purpose locally differentially-private synthetic data release via spline wavelet plug-in estimation
Thibault Randrianarisoa, Lukas Steinberger, Botond Szabó
We develop plug-in estimators for locally differentially private semi-parametric estimation via spline wavelets. The approach leads to optimal rates of convergence for a large clas…
Uncertainty quantification via cross-validation and its variants under algorithmic stability
Nicolai Amann, Hannes Leeb, Lukas Steinberger
Recently, there has been substantial interest in statistical guarantees for cross-validation (CV) methods of uncertainty quantification in statistical learning (cf. Barber et al. 2…
Efficient Estimation of a Gaussian Mean with Local Differential Privacy
Nikita P. Kalinin, Lukas Steinberger
In this paper we study the problem of estimating the unknown mean of a unit variance Gaussian distribution in a locally differentially private (LDP) way. In the high-privacy r…