Learning Theory of Classification with Deep Neural Networks

Abstract: Deep neural networks have achieved remarkable success in various binary classification tasks. Despite their practical effectiveness, theoretical understanding of their generalization in binary classification remains limited. In this talk, I will present our recent progress on classification using deep neural networks.

CV: 石磊，复旦大学数学科学学院教授，博士生导师。研究方向为学习理论和逼近论，主要研究机器学习算法的逼近，泛化和优化理论。相关成果发表于Applied and Computational Harmonic Analysis、Foundations of Computational Mathematics、Inverse Problems、Annals of Statistics、Mathematical Programming、SIAM Journal on Optimization以及Journal of Machine Learning Research等应用数学，统计，优化以及机器学习领域的国际权威期刊。主持上海市基础研究重点项目以及国自然面上项目，参与国自然重点项目以及中港联合基金项目，入选上海市优秀学科带头人计划。