The 2023 International Workshop on Learning and Information Theory (WOLIT'23) was held at the Shenzhen Institute for Talents Development and Hong Kong Charles K. Kao Auditorium from July 3 to 5. It was jointly hosted by the Institute of Data and Information of Tsinghua Shenzhen International Graduate School (Tsinghua SIGS), the Institute of Network Coding of The Chinese University of Hong Kong (CUHK), and the Shenzhen Research Institute of Big Data.
Group photo at Hong Kong venue
Group photo at Shenzhen venue
The workshop featured eighteen distinguished scholars from prestigious universities and research institutes, such as the Massachusetts Institute of Technology, Columbia University, the University of Hong Kong,and The Chinese University of Hong Kong (CUHK). Scholars shared their latest research findings on learning theory and information theory with nearly 200 on-site participants. Faculty and students from institutions including Tsinghua SIGS, the University of Hong Kong, The Chinese University of Hong Kong at Shenzhen (CUHK-Shenzhen), and the Southern University of Science and Technology (SUSTech) attended the on-site workshop and participated in the Student Poster Showcase. Livestreams of the three-day event attracted over 3,400 viewers on Chaspark and WeChat Video.
Xiao-Ping Zhang delivering opening remarks
General Chair of WOLIT’23, Shao-Lun Huang, a tenured associate professor at Tsinghua SIGS, moderated the opening ceremony. Opening speeches were delivered by Xiao-Ping Zhang, fellow of the Canadian Academy of Engineering, fellow of IEEE, dean of the Tsinghua SIGS Institute of Data and Information, and Tsinghua Pengrui Professor, Raymond W. Yeung, Choh-Ming Li Professor of Information Engineering and co-director of the Institute of Network Coding at CUHK, Zhi-Quan Luo, vice President of CUHK-Shenzhen and director of the Shenzhen Research Institute of Big Data, and Yi Ma, director and chair professor of HKU Musketeers Foundation Institute of Data Science and head of the CUHK Department of Computer Science.
Raymond W. Yeung, Yi Ma, Masahito Hayashi, Rui Zhang, Ruoyu Sun, and Ercan E. Kuruoglu (from left to right, top to bottom) give keynote talks
On the morning of July 3, opening speakers, including Raymond W. Yeung, Yi Ma, and Masahito Hayashi, discussed the development of machine-proving entropy inequality, a systematic view of artificial intelligence based on parsimonious and structured representations, and the generalization of the Arimoto-Blahut algorithm in machine learning and information theory, respectively. On the afternoon of July 3, speakers including Rui Zhang, Ruoyu Sun, and Ercan E. Kuruoglu presented the latest advancements on intelligent reflecting surfaces (IRS) in wireless networks, the theoretical properties of the Adam optimizer, and time-varying topological signal processing.
Panel discussion
In the subsequent panel discussion section, panelists Xiao-Ping Zhang, Raymond W. Yeung, Lizhong Zheng, and Chao Tian answered key questions related to information theory research and innovation in the era of “AI explosion.”
Lizhong Zheng, Chao Tian, Amin Gohari, Guangyue Han, and Haim Permuter (from left to right, top to bottom) give keynote talks
On the morning of July 4, Lizhong Zheng, Chao Tian, and Guangyue Han presented their latest research on neural network-based modal decomposition, information theoretic generalization error bound for learning algorithms, and the sampling problem for continuous-time additive Gaussian white noise channels with possible feedback. In the afternoon session, speakers Haim Permuter, Chandra Nair, and Amin Gohari discussed topics related to information measurement in network channels, non-convex optimization problems in information theory, and the correlation between Shannon-type inequalities and generalized relative entropy.
Student poster presentation session
Afterwards, during the Student Poster Showcase, thirty-six students from Tsinghua SIGS, the University of Hong Kong, CUHK-Shenzhen, and SUSTech presented their posters to the workshop participants. Anping Zhang from Tsinghua SIGS, Fan Liu from SUSTech, and Chen Li and Yongheng Wang from Tsinghua SIGS were awarded first, second, and third place, respectively, in the student poster competition.
Deniz Gunduz, Cynthia Rush, Nikolaos Pappas, Uri Cohen, Photios Stavrou, and Yihan Zhang (from left to right, top to bottom) give keynote talks
On the morning of July 5, Deniz Gunduz, Cynthia Rush, and Nikolaos Pappas discussed topics related to designing communication systems for modern applications, L1 regularization in high-dimensional statistical estimation problems, and meeting the requirements of emerging IoT applications. During the afternoon session, speakers Uri Cohen, Photios Stavrou, and Yihan Zhang talked about using statistical physics to measure the classification ability of deep neural networks, the application of rate-distortion theory in target-oriented semantic communication, and extracting information components from noise using spiked random matrices.
Source: Institute of Data and Information
Written by Chen Yanyu, Chen Bohan & Guo Jia
Edited by Alena Shish & Yuan Yang