题目：Co-Clustering of Nonsmooth Graphons

主讲：David Choi教授美国卡内基梅隆大学公众政策与信息管理系

时间：7月7号（周四）上午9：00开始

地点：管理学院314教室

Abstract: Theoretical results are becoming known for community detection and clustering of networks; 

however, these results assume an idealized generative model that is unlikely to hold in many settings. 

Here we consider exploratory co-clustering of a bipartite graph, where the rows and columns of 

the adjacency matrix are assumed to be samples from an arbitrary population. This is equivalent to 

assuming that the data is generated from a nonparametric model known as a graphon. We show that 

co-clusters found by any method can be extended to the row and column populations, or equivalently 

that the estimated blockmodel approximates a blocked version of the generative graphon, with generalization

 error bounded by n^{-1/2}. Analogous results are also shown for degree-corrected co-blockmodels 

and random dot product bipartite graphs, with error rates depending on the dimensionality of the latent 

variable space.

个人简历：

David Choi教授于2004年毕业于斯坦福大学电子信息工程系获得博士学位。后在美国麻省理工学院Lincoln实验室做研究员。2009-2011在哈佛大学工业工程系做博士后研究。2011-2012年在加州大学伯克利分校统计系做兼职教授。2012年进入卡内基梅隆大学公共政策与信息管理系任教。