BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Identification and Estimation of Graphical Continuous Lyapunov Mod
 els - Mathias Drton (Technical University of Munich)
DTSTART:20250207T140000Z
DTEND:20250207T150000Z
UID:TALK226141@talks.cam.ac.uk
CONTACT:Qingyuan Zhao
DESCRIPTION:Graphical continuous Lyapunov models offer a new perspective o
 n modeling causally interpretable dependence structure in multivariate dat
 a by treating each independent observation as a one-time cross-sectional s
 napshot of a temporal process. Specifically\, the models consider multivar
 iate Ornstein-Uhlenbeck processes in equilibrium. This leads to Gaussian m
 odels in which the covariance matrix is determined by the continuous Lyapu
 nov equation. In this setting\, each graphical model assumes a sparse drif
 t matrix with support defined by a directed graph. The talk will discuss t
 he identifiability of such sparse drift matrices and their regularized est
 imation.
LOCATION:Centre for Mathematical Sciences MR15\, CMS
END:VEVENT
END:VCALENDAR