1: \begin{abstract}

2: In this article, we consider the residual regularization path-following method with

3: the trust-region updating strategy for the linear complementarity problem. This 

4: time-stepping selection based on the trust-region updating strategy overcomes 

5: the shortcoming of the line search method, which consumes the unnecessary trial 

6: steps in the transient-state phase. In order to improve the robustness of the 

7: path-following method, we use the residual regularization parameter to replace 

8: the traditional complementarity regularization parameter. Moreover, we prove the 

9: global convergence of the new method under the standard assumptions without the 

10: traditional assumption condition of the priority to feasibility over complementarity. 

11: Numerical results show that the new method is robust and efficient for the linear 

12: complementarity problem, especially for the dense cases. And it is more robust 

13: and faster than some state-of-the-art solvers such as the built-in subroutines 

14: PATH and MILES of the GAMS v28.2 (2019) environment. The computational time of 

15: the new method is about 1/3 to 1/10 of that of PATH for the dense linear 

16: complementarity problem.

17: \end{abstract}

18: