Solving a class of nonconvex quadratic programs by inertial DC algorithms
Người báo cáo: Nguyễn Năng Thiều
Thời gian: 9h00 đến 11h00 sáng thứ Tư ngày 05.11.2025.
Địa điểm: Phòng 508 nhà A6 Viện Toán học
Tóm tắt: In this talk, we present two inertial DC algorithms for indefinite quadratic programs under linear constraints (IQPs), where the constraint set may be unbounded. Using a qualification condition involving the normal cones of unbounded pseudo-faces of the polyhedral convex constraint set, the recession cones of the corresponding faces, and the quadratic form defining the objective function, we show that the resulting iteration sequences are bounded whenever the given IQP has a finite optimal value. Any cluster point of such a sequence is a Karush-Kuhn-Tucker point. Moreover, all cluster points of a given iteration sequence belong to a single connected component of the Karush-Kuhn-Tucker point set.
