A rapid algorithm for a class of linear complementarity problems (Q2372045)

The paper deals with the linear complementarity problem: Find \(x\in\mathbb{R}^n\) such that \(x\geq 0\), \(F(x)\geq 0\), \(F(x)_i x_i= 0\), where \(F(x)= Mx+ q\). Usually, this problem is approximately solved by a projection method. The projection method is globally convergent under suitable assumptions, but this convergence is slow. \textit{O. L. Mangasarian} [J. Optimization Theory Appl. 22, 465--485 (1977; Zbl 0341.65049)] presented a better method which is described in the present paper. The authors introduce a new projection method based on matrix split skill and proximal point algorithms. It is an implicit method. A modified algorithm possesses the advantage of robust convergence. Let \(M\) be symmetric and positive definite, then the sequence generated by the new algorithm converges and the limit solves the linear complementarity problem. Numerical results obtained for three examples illustrate the new method. They show that the algorithm is almost not sensitive to the parameters.