学术报告（王军辉 2025.3.20）

摘要：Network data has attracted increasing research interests across various scientific communities. In this talk, we will talk about some recent development of network embedding model, including directed network and hypergraph network. Interestingly, network embedding model resembles latent factor model in classical statistics, which can be dated back to early 1900s. Along this line, we will develop some new modeling strategies in the same spirit of embedding unstructured network into well-structured vector sapce, to tackle various issues arising from these applications. The developed methods are supported by their asymptotic properties as well as a variety of real applications. If time permits, I will also briefly discuss about structural change point detection via embedding dynamic networks.