Theoretical study on deep learning: approximation,generalization and generation

Release time:2023-05-08Views:243

Title: Theoretical study on deep learning: approximation, generalization and generation


Speaker: Yuling JiaoWuhan University

Time:Friday, May 19, 2023, 15:30-16:30

Location: Zoom ID:876 0592 8254 , Password:2023 

Abstract: In the first part of this talk, I will discuss some theoretical studies on deeplearning with a focus on approximation, generalization. In particular, I willcover error analysis with over-parameterization. In the second part, I will delveinto the concept of Gaussian stochastic interpolations and their applications,including the derivation of functional inequality with dimension-free constantsas well as sampling and generative learning.

Biography: Dr. Yuling Jiao is an Associate Professor in the School of Mathematicsand Statistics in Wuhan University. His research interests include machineearning, scientific computing. His research works were published in the Annalsof Statistics, Journal of the American Statistical Association, Statistical Science!A Review Journal ofIMS, SIAM Journal on Mathematical Analysis, SIAMournal on Numerical Analysis, SIAM Journal on Scientific Computing, Appliecand Computational Harmonic Analysis. Inverse Problems, IEEE Transactions onSignal Processing. Journal of Machine Learning Research, ICML, NeurIPS.

More information about HIT-WHU Seminar on Stochastic Analysis and Algorithms can be found here.

Copyright (C)2017 Institute for Advanced Study in Mathematics of HIT All Rights Reserved.
Contact Us:
Add:NO.92 West Da Zhi St. Harbin China
Technical support:Net & Information Center,HIT