【主讲】Yichong Zhang (教授，Singapore Management University)
【主题】Determining the Number of Communities in Degree-corrected Stochastic Block Models
【时间】2018年5月25日 (周五) 15：30-17：00
【地点】上海财经大学经济学院楼701室
【语言】英文
【摘要】In this paper, we propose to estimate the number of communities in degree-corrected stochastic block models based on a pseudo likelihood ratio in which the community memberships are computed by spectral clustering. We show that our estimator of the number of communities is consistent even when the average degree of the network grows at a polylog rate. In simulations, we compare our method with existing methods in the literature and demonstrate its satisfying performance in scenarios that the network is sparse and/or has unbalanced communities.
