题　目：Taming the data deluge

嘉　宾：黄增峰，教授，复旦大学

主持人：林学民，教授，上海交通大学安泰经济与管理学院

时　间：2024年1月12日（周五）13:30-15:00

地　点：安泰经济与管理学院A407

内容简介: There has been a notable upswing in the amount of data collected and processed in modern applications. With the generation of more data than can be effectively and efficiently managed or constrained, there are missed opportunities for the analysis and interpretation of data to inform decision-making.  Handling such massive data surpasses the capacities of traditional algorithmic theories. In particular, traditional theory used simplistic computation models which are not suitable for modern computer architectures and applications.

In this talk, I will introduce modern techniques to address this issue. First I will focus on data summarization, which builds a compact summary of a dataset that is vastly smaller but well approximates the original data. I will provide an overview and illustrate how to design and analyze efficient summarization algorithms in big data computation models. Next, I will discuss scalability issues of machine learning methods on graph data and provide techniques that effectively overcome the challenges. Finally, I will talk about big data related problems in the setting of online learning.

演讲人简介：黄增峰，复旦大学大数据学院教授，博士生导师。浙江大学和香港科技大学分别获得学士和博士学位。在国际期刊和会议上（ICML, NeurIPS, JMLR, FOCS, SICOMP, SIGMOD, TIT等）发表高水平论文五十余篇。获上海市扬帆计划、国家级青年人才计划支持。所发表论文荣获了多项国际奖项，包括ICML 2018最佳论文亚军奖（best paper runner up award）， 两次世界人工智能大会青年优秀论文提名奖（2020，2023），以及PODS 2022时间检验奖。

欢迎广大师生参加！