论坛时间：2021年12月21日8:00-9:30；14:00-15:30
论坛形式：腾讯会议，ID 266 826 250
报告及报告人简介
报告1:Data Reduction for Communication-efficient Machine Learning
报告人：Ting He
报告人单位：宾夕法尼亚州立大学
报告时间：2021年12月21日8:00-9:30
报告人简介：
Dr. Ting He is an Associate Professor in the Department of Computer Science and Engineering at the Pennsylvania State University, University Park, PA. Her interests include computer networking, performance evaluation, statistical inference, and machine learning. She has served as an Associate Editor for IEEE Transactions on Communications and IEEE/ACM Transactions on Networking, a TPC Co-Chair of IEEE ICCCN, an Area TPC Chair of IEEE INFOCOM, and a TPC member of many international conferences in the areas of networking and distributed computing. Her work has been recognized by many awards, including the Research Division Award and multiple Outstanding Contributor Awards from IBM, the Military Impact, the Commercial Prosperity, and the Most Collaboratively Complete Publications Awards from ITA, the 2021 IEEE Communications Society Leonard G. Abraham Prize, the Best Paper Award at the 2013 ICDCS, the Outstanding Student Paper Award at the 2015 SIGMETRICS, the Best Student Paper Award at the 2005 ICASSP, and multiple best paper finalists. She is an active contributor in N2Women and was listed in“N2Women: Rising Stars in Networking and Communications”in 2017.
报告简介：
The rapid growth of smart phones and IoT devices leads to a huge amount of data generated at the edge of communication networks. While machine learning has provided powerful tools to exploit the information in such big data, these tools tend to be resource-intensive, making them unsuitable for the edge devices. Meanwhile, due to bandwidth and other constraints, it is also often infeasible to transfer all the data to servers that have the required computation resources. In this talk, we will talk about how to address this challenge through various data reduction techniques, with focus on the construction of smaller weighted datasets, known as coresets, that can be used as proxies of the full dataset. This work has won the Military Impact and the Commercial Prosperity Awards from the International Technology Alliance in Distributed Analytics and Information Sciences (DAIS-ITA).
报告2:面向AIOT应用的边缘智能优化技术
报告人：东方
报告人单位：东南大学
报告时间：2021年12月21日14:00-15:30
报告人简介：
东方，博士，东南大学青年特聘教授，博士生导师。现任东南大学大数据计算中心主任，计算机科学与工程学院计算机工程系主任。同时担任ACM南京分会共同主席、ACM SIGCOMM China秘书长。主要研究方向为云计算与边缘计算。主持国家重点研发计划项目课题/子课题2项、国家自然科学基金3项。参加了诺贝尔物理学奖获得者丁肇中教授领导的AMS大型物理实验，建设完成东南大学大数据中心及东南大学AMS科学数据处理中心（AMS-02 SOC）。在INFOCOM、ICDCS、TMC、TSC等重要期刊及会议上发表论文70余篇。
报告简介：
随着深度学习技术的发展以及智能终端设备的普及，人工智能和物联网的结合已成为必然趋势，越来越多的智联网（AIOT）应用场景应运而生。现有计算模式无法满足AIOT应用高精度低延时的需求。因此需要设计一种新型的计算框架及相关优化机制，实现AIOT应用的实时高效执行。本报告主要通过对现有深度学习加速机制与应用部署模式进行分析，引出边缘计算和边缘智能的概念，并探讨边缘智能计算框架与相关的执行优化技术。
