數學系 Seminar 第 2040 期
報告主題：Iterative algorithm for solving a class of convex feasibility problem
報告人：段雪峰 教授 （桂林電子科技大學）
報告摘要：In this talk, we consider a class of convex feasibility problem, which arises in quantum computation. Based on the matrix equation theory, the feasible sets are characterized by exploiting the special structure of the linear constraints, and its analytic expression is given. By making use of the nice structure properties and the KKT condition, we derive the projection formulas of a matrix onto the feasible sets. The relaxed alternating projection method is designed to solve the convex feasibility problem. Numerical experiments show that the new method is feasible and effective.