Workshop on Numerical Linear Algebra and Optimization

In recent years, optimization problems involving eigenvalues have received a lot of attention. One well-studied class of convex problems is semidefinite programming, which has nonnegativity constraints placed on eigenvalues. Additionally, many important nonconvex problems involve constraints on the eigenvalues of nonsymmetric matrices, or their generalization, namely pseudospectra. These problems possess intrinsic similarities, yet they are studied by different communities. The workshop aims to bring these communities together, as well as to share modern applications of these problems in fields such as machine learning and control. Effective use of large-scale numerical linear algebra tools will also be a theme of the workshop.

Michael Overton is well-known for his seminal work in optimization and numerical linear algebra, and the selected topics are representative of his research interests. The meeting will represent an opportunity to recognize his many contributions, on the happy occasion of his 60th birthday.