Academic lecture of College of Information Science and Technology / College of Cyber Security(No.32)
Topic: Collaborative filtering with awareness of social network
Lecture | October17, 2018 |10:00a.m. | Room 224, Nahai Building
Speaker: Jing Bing-yi, The Hong Kong University of Science and Technology(HKUST)

Abstract:

We propose the so-called NetRec method in recommender system by incorporating the network information into collaborative filtering (CF). This results in a sharper error bound than previous literature under reasonable assumptions. It is also shown that the combination of the network-related penalty and the nuclear norm penalty gives better estimates than those achieved by any of them alone. The method has been shown to work well in simulations and some real data sets on Yelp. This is a joint work with X.S. Yu, T. Li, and N.C. Ning.

ABOUT JING BING-YI:

Prof. Jing received his PhD in Statistics at the University of Sydney in 1993. He is currently a professor at the Department of Mathematics and the director of the Center for Statistical Science in HKUST. He is also a part-time professor of the Yangtze River Scholar Program and a part-time professor of Jinan University. He is an elected member of the International Statistical Institute (ISI) and the Board Director of ICSA. He is serving as Associate Editors for five journals, including Journal of Business & Economic Statistics, Canadian Journal of Statistics. In 2010, he won the second class of Higher Education Outstanding Scientific Research Output Awards (Natural Science) and the second class of State Natural Science Award in 2015. He has wide research interests, including probability and statistics, bioinformatics, financial econometrics, machine learning and data mining. So far, he has published over 100 research papers, including more than 20 papers in high-level journals, like The Annals of Statistics,The Annals of Probability and Journal of the American Statistical Association.