Wednesday, September 27

Minisymposium 1: Linear algebra issues in optimization methods

Time: 10:30 - 12:30
Room: L1.202, Building L
Organiser: Martin Stoll, MPI Magdeburg, and Margherita Porcelli, Università degli Studi di Firenze
Chair: Martin Stoll, MPI Magdeburg

Computational mathematicians dealing with simulations of large-scale optimization problems based on physical phenomena have made tremendous success over the last decades. This has enabled scientists from various areas of engineering, chemistry, geophysics, et al. to ask more relevant and complex questions. With the consideration of more sophisticated models efficient optimization methods required efficient techniques from numerical linear algebra to enable or speed up the solution process. In this minisymposium we present some of the latest developments for the fast and robust solution of linear algebra problems within optimization methods of various types.

 

Speakers:

10:30 - 11:00Daniela di Serafino (University of Campania Luigi Vanvitelli)
Constraint-preconditioned Krylov methods for regularized saddle-point systems
11:00 - 11:30John Pearson (University of Kent)
Linear algebra for time-dependent PDE-constrained optimization
11:30 - 12:00Lena Vestweber (TU Braunschweig)
Douglas-Rachford iterations for two and three dimensional TV-, TGV- and constrained TGV denoising
12:00 - 12:30Justin Buhendwa Nyenyezi (University of Namur)
Preconditioning linear systems from deformable 3D medical image registration using low-rank tensor train format