报告题目：Towards Robust Optimization: Three Different Random Sampling Approaches
In many real-world scenarios, our datasets often contain significant outliers. The outliers can be naturally generated in the data collection process, or added by some adversarial attacker. For example, an attacker can inject a small number of specially crafted samples into the training data which make the decision boundary severely deviate and cause unexpected misclassification (this event is called a poisoning attack). Therefore, designing robust optimization algorithms, in particular being resilient to outliers, has become a popular topic and attracted a great amount of attention in recent years. However, most robust optimization problems have very high complexities and a number of recent research focus on how to reduce their complexities and how to speed up their corresponding algorithms. Random sampling is a natural idea for reducing data size, but existing sampling methods are often difficult to be extended to handle the cases involving outliers. In this talk, we will introduce three different novel random sampling approaches for handling several popular robust optimization problems in the fields of machine learning and data mining. Some of the mentioned results have been published in ICML’20, ESA’19, and ESA’20.
Hu Ding is a (pre-tenure) professor in The School of Computer Science and Engineering at USTC, and direct the Data Intelligence, Algorithms, and Geometry (DIAG) group. Before moving back to China, I was a tenure-track assistant professor in the department of computer science and engineering at Michigan State University for a short time (2016-2018). I held a joint research fellow position of Tsinghua University and UC Berkeley from 2015 to 2016, which is titled as "Simons-Berkeley Research Fellow". I got my Ph.D under supervision of Dr.Jinhui Xu, in the Department of Computer Science and Engineering, State University of New York at Buffalo, in 2015. I received my bachelor degree in Mathematics from Sun Yat-Sen (Zhong Shan) University in 2009. My research interests lie in the fields of Algorithms, Computational Geometry, and their applications in real world, e.g., Machine Learning, Big Data, Internet of Things, Computer Vision, and Biomedical Imaging. Homepage: http://staff.ustc.edu.cn/~huding/index.html