Xin (Henry) Zhang

Department of Statistics, Florida State University

Title: Envelopes in Discriminant Analysis and Model-based Clustering


In this talk, we will discuss the applications of envelopes to various supervised and unsupervised learning problems. In particular, we will discuss on how to improve estimation and prediction in classification problems with vector and tensor data, and how to efficiently identify latent clusters and heterogeneity in the unlabeled data.