首页-->科学研究-->教学科研动态

IT讲坛:Low-dimensional Modeling for Deep Learning

时间:2021-12-24 09:22:17 文章来源 :学科 浏览量:147


会议主题:Low-dimensional Modeling for Deep Learning

会议时间:2021/12/28 09:30-12:00 (GMT+08:00) 中国标准时间 - 北京

点击链接入会,或添加至会议列表:

https://meeting.tencent.com/dm/pWZW5PZmaqsn

#腾讯会议:942-219-480


Abstract: In the past decade, the revival of deep neural networks has led to dramatic success in numerous applications ranging from computer vision to natural language processing, to scientific discovery and beyond. Nevertheless, the practice of deep networks has been shrouded with mystery as our theoretical understanding for the success of deep learning remains elusive.

In this talk, we will exploit low-dimensional modeling to help understand and improve the performance of deep learning. In particular, we will focus on the following two perspectives: (i) developing principled approach for robust recovery of natural images by over-parameterizing images with deep convolutional networks, and (ii) understanding neural collapse which implies that the representations by a deep learning classifier span a very low-dimensional space, an intriguing empirical phenomenon that persists across different neural network architectures and a variety of standard datasets. We will exploit our understanding of neural collapse to improve training efficiency.

1

Bio: Zhihui Zhu is currently an Assistant Professor with the Department of Electrical and Computer Engineering, University of Denver, CO, USA. He was a Post-Doctoral Fellow with the Mathematical Institute for Data Science, Johns Hopkins University, from 2018 to 2019. He received his B.Eng. degree in communications engineering in 2012 from Zhejiang University of Technology (Jianxing Honors College), and his Ph.D. degree in electrical engineering in 2017 at the Colorado School of Mines, where his research was awarded a Graduate Research Award. His research interests include the exploitation of inherent low-dimensional structures within data and signals, and the design, analysis, and implementation of optimization algorithms for machine learning and signal processing.