2015年12月24日，美国加州大学戴维斯分校教授Cho-Jui Hsieh应电子工程与信息科学系和多媒体计算与通信-教育部微软重点实验室的邀请在科大西区科技实验楼西楼1213会议室做题为《Computational and Statistical Challenges in Matrix Completion》的高水平学术前沿讲座。该讲座中，Cho-Jui Hsieh教授介绍了其近年来在大规模机器学习优化算法和技术方面的研究成果，并与参会师生进行了升入的交流和探讨。 

　　图1 报告人：Cho-Jui Hsieh (University of California, Davis, USA) 

　　图2 Cho-Jui Hsieh做题为《Computational and Statistical Challenges in Matrix Completion》的报告

报告摘要：
Matrix completion has become popular for building large-scale recommender systems. In this talk, we discuss the computational and statistical challenges when we apply matrix completion algorithms to big data applications. In the first part of the talk, we describe a non-locking stochastic multi-machine algorithm for Asynchronous matrix completion (NOMAD), and conduct experiments on a dataset with more than 2 billion ratings. In the second part of the talk, we describe a technique to reduce the sample complexity of matrix completion algorithms by incorporating noisy side information. This is joint work with Hsiang-Fu Yu, Kai-Yang Chiang, Inderjit S. Dhillon, S.V.N. Vishwanathan and Hyokun Yun.

报告人简介：
Cho-Jui Hsieh is an assistant professor of department of computer science and statistics at University of California, Davis. His research focus is developing new algorithms and optimization techniques for large-scale machine learning problems. Cho-Jui obtained his M.S. degree in 2009 from National Taiwan University (advisor: Chih-Jen Lin) and Ph.D. from University of Texas at Austin in 2015 (advisor: Inderjit S. Dhillon). He is the recipient of the IBM Ph.D. fellowship in 2013-2015, the best research paper award in KDD 2010, and the best paper award in ICDM 2012.