An APPA-based descent method with optimal step-sizes for monotone variational inequalities (Q2462142)

The goal of this paper is to present the new approximate proximal point algorithms (APPA)-based descent method, to solve monotone variational inequalities. In general the paper is concentrated on improving some existing APPA-based descent methods by choosing optimal step-sizes. The numerical results demonstrate that the selection of optimal step-sizes in both the extra-gradient step and the descent step reduces considerable computational load of the APPA-based descent methods. Main result: The superiority to some existing methods is verified both theoretically and computationally. Global convergence of the new method is proved under mild assumptions. Moreover the new method is compared to the HYY-method of \textit{B.-S. He, Z.-H. Yang} and \textit{X.-M. Yuen} [J. Math. Anal. Appl. 300, No.~2, 362--374 (2004; Zbl 1068.65087)], whose theoretical assertion is verified by the proposed numerical results.