Events Calendar

AI SEMINAR

Speaker: Xiaoran Yan, University of New Mexico

Talk Title: Variational Inference of Community Models: A Unifying Learning Framework

Series: AISeminar

Abstract: Community detection is an important part of network modelling, as community structure offers clues to the processes which generated the networks. We propose a general variational framework for learning community based network models, which offers a range of approximation options. In particular, a message passing algorithm under this framework achieves a good balance between accuracy and scalability. The framework is closely related to many other popular algorithms for community detection, including random walk and spectral clustering. To showcase an application, we will study the model selection problem for stochastic block models, in which the framework is adapted to learn information theoretic measures for comparing candidate models.

Biography: Xiaoran Yan is a computer science Ph.D. candidate at the University of New Mexico. He also works as a graduate fellow at the Santa Fe Institute under the advisement of Cris Moore. His work focus on community detection on complex networks, including building statistical models and developing scalable learning algorithms. He has recently defended his dissertation on model selection for stochastic block models. He got his bachelor's degree at the Zhejiang University in China.

Host: Kristina Lerman

Webcast: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=8013c24dbbde4b42ba0f10da844fe2621d

Location: Information Science Institute (ISI) -

WebCast Link: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=8013c24dbbde4b42ba0f10da844fe2621d

Audiences: Everyone Is Invited

Contact: Alma Nava / Information Sciences Institute