Extremal graphical lasso and high-dimensional extremes

The speaker is Dr. Sebastian Engelke (University of Geneva)

Statistical inference for the extremal graphical models introduced in Engelke and Hitz (2020, JRSSB) is so far restricted to simple structures called block graphs. We develop an extremal graphical lasso that can be used to estimate in a data-driven way the structure in general Huesler-Reiss graphical models. We propose an efficient algorithm and prove that it recovers the underlying graph structure consistently even for growing dimension. This enables the use of the extremal graphical lasso in high-dimensional settings. 

This seminar is taking place on Zoom. If you are interested in joining us, please contact Lanpeng Ji (l.ji@leeds.ac.uk) or Konstantinos Dareiotis (k.dareiotis@leeds.ac.uk) for Zoom detail.