Seminar | Institute of Mathematical Sciences
Time：16:00-17:00, June 28, Friday
Location：Room S407, IMS
Speaker: Jie Peng, Department of Statistics, University of California, Davis
Abstract: In this talk, we consider learning latent graphs from multivariate (stochastic) observations where each variable is identified with a node of the graph and the graph topology reflects certain characteristics of the graph-referenced signals. We will consider various models and discuss their theoretical and computational aspects. The discussion will focus on the high dimensional regime where the number of variables (nodes) can be much larger than the sample size.