Second-order cone and semidefinite methods for the bisymmetric matrix approximation problem

Abstract Approximating the closest positive semi-definite bisymmetric matrix using Frobenius norm to a data is important in many engineering applications, communication theory and quantum physics. In this paper, we will use interior point method solve problem. The problem be reformulated into various forms, beginning as programming later, form of mixed semidefintie second-order cone optimization Numerical results comparing efficiency these methods with alternating projection algorithm reported.