Abstract

This session focuses on modern statistical methods for multivariate and structured data arising in a wide range of scientific applications. Topics include low-rank and latent variable models, dimension reduction, matrix completion, dependence modeling, multivariate prediction, and the incorporation of auxiliary covariate information. The session will bring together recent advances in statistical learning, machine learning, and stochastic modeling for high-dimensional data, including flexible approaches for modeling complex relationships between multivariate responses and explanatory variables. The goal is to highlight both methodological developments and emerging applications in the analysis of modern multivariate data.