计算与应用讨论班

报告题目：On Algorithmic Stability and Robustness of Bootstrap SGD

报告人：Andreas Christmann（University of Bayreuth, Germany）

时间：2026年05月21日（星期四），15:00

地点：海纳苑2幢204

摘要：The bootstrap is a computer-based resampling method that can provide good approximations to the finite sample distribution of a given statistic. In this talk some methods to use the empirical bootstrap approach for stochastic gradient descent (SGD) to minimize the empirical risk over a Hilbert space are investigated from the view point of algorithmic stability and statistical robustness. Two types of approaches are based on averages and are investigated from a theoretical point of view. Another type of bootstrap SGD is proposed to demonstrate that it is possible to construct purely distribution-free pointwise confidence intervals and distribution-free pointwise tolerance intervals of the conditional median function using bootstrap SGD.

报告人简介：Prof. Christmann got his dissertation and habilitation at the University of Dortmund (Germany).After positions as a visiting professor at KU Leuven (Belgium) and as professor at universities in Dortmund (Germany) and Brussels (VUB, Belgium).He is serving as Full Professor and Chair of Stochastics and Machine Learning at the University of Bayreuth (Germany) since 2008.Together with Prof. Steve Smale, Prof. Ding-Xuan Zhou, and Prof. Kurt Jetter, he organized the Oberwolfach workshop Learning Theory and Approximation, July 3-9, 2016.He was Action Editor of Journal of Machine Learning Research (JMLR) from 2013 to 2019 and since 2020 he is amember of the JMLR Editorial board of reviewers.Together with Prof. Ingo Steinwart he published a Springer book on Support Vector Machines.His main research topics are statistical learning theory and robust statistics.

联系人：郭正初（guozc@zju.edu.cn）