Mathematical optimization and statistical theories using geometric methods

Information

Date: October 20--21, 2022 (Japan Standard Time)
Venue: Academic Extension Center (Osaka Metropolitan University)
Contents: Workshop (Hybrid: physical/virtual)

This workshop has ended successfully.

Program

• October 20 (Thurseday)

13:00--13:50 Shoji Toyota (SOKENDAI)
Invariance Learning based on Label Hierarchy
14:00--14:50 Sho Sonoda (RIKEN AIP)
Ridgelet Transforms for Neural Networks on Manifolds and Hilbert Spaces
15:00--15:50 Tomonari Sei (The University of Tokyo)
Ushio Tanaka (Osaka Metropolitan University)
Stein-type distributions on Riemannian manifolds
16:10--17:00 Tomasz Skalski (Wroclaw University of Science and Technology: LAREMA, University of Angers)
On LASSO and SLOPE estimators and their pattern recovery
17:10--18:00 Carlos Améndola (Technical University of Berlin)
Likelihood geometry of correlation models

• October 21 (Friday)

9:00-- 9:50 Piotr Zwiernik (University of Toronto)
Mixed convex exponential families and locally associated graphical models
11:00--11:50 Koichi Tojo (RIKEN Center for Advanced Intelligence Project)
Classification problem of invariant q-exponential families on homogeneous spaces
13:50--14:40 Yoshihiko Konno (Osaka Metropolitan University)
Adaptive shrinkage of singular values for a low-rank matrix mean when a covariance matrix is unknown
14:50--15:40 Satoshi Kuriki (The Institute of Statistical Mathematics)
Expected Euler characteristic heuristic for smooth Gaussian random fields with inhomogeneous marginals
16:00--16:50 Piotr Graczyk (LAREMA, University of Angers)
Pattern recovery by SLOPE

Organizers

Contact: hideto [at] ism.ac.jp

Sponsor

This workshop is supported by the following institution and grant:

Last modified: October 22, 2022