Parameterized error bounds and their optimality for linear complementarity problems involving Nekrasov and B-Nekrasov matrices
摘要
This paper develops new parameterized error bounds for linear complementarity problems (LCPs) involving Nekrasov matrices and B-Nekrasov matrices. By utilizing the monotonicity properties of the parameter-dependent functions, we provide a complete characterization of the optimal values of these error bounds. Our analysis demonstrates that, under appropriate conditions, the proposed optimal bounds strictly improve those established by Dai et al. (Numer. Math. Theor. Meth. Appl., 12(4): 1191-1212, 2019). Supporting numerical experiments validate the theoretical results and highlight the efficacy of the new bounds.