Abstract
AbstractApproximating the closest positive semi-definite bisymmetric matrix using the Frobenius norm to a data matrix is important in many engineering applications, communication theory and quantum physics. In this paper, we will use the interior point method to solve this problem. The problem will be reformulated into various forms, in the beginning as a semi-definite programming problem and later, into the form of a mixed semidefintie and second-order cone optimization problem. Numerical results comparing the efficiency of these methods with the alternating projection algorithm will be reported.
Publisher
Springer Science and Business Media LLC
Reference21 articles.
1. Aitken, A.C.: Note on a special persymmetric determinant. Ann. Math. 2(32), 461–462 (1931)
2. Al-Homidan, S.: Low rank methods for solving the nearest correlation matrix problem. J. Nonlinear Convex Anal. 19(6), 881–892 (2018)
3. Al-Homidan, S.: Structure method for solving the nearest Euclidean distance matrix problem. J. Inequal. Appl. 1, 491 (2014)
4. Al-Homidan, S.: Approximate Toeplitz problem using semidefinite programming. J. Optim. Theory Appl. 135(3), 583–598 (2007)
5. Alizadeh, F.; Haeberly, J.; Overton, M.: Primal-dual interior-point methods for semi-definite programming: convergence rates, stability and numerical results. SIAM J. Optim. 8, 746–768 (1998)
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献