Abstract:A new error bound for the linear complementarity problem is given when the involved matrix is a B-matrix. It is shown that this bound is sharper than some previous bounds [C.Q. Li, Y.T. Li. Note on error bounds for linear complementarity problems for B-matrices, Applied Mathematics Letters, 57:108-113,2016] and [C.Q. Li, Y.T. Li. Weakly chained diagonally dominant B-matrices and error bounds for linear complementarity problems, to appear in Numer.Algor.].
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to provide a new error bound for the Linear Complementarity Problem (LCP), especially when the matrix involved is a B - matrix. Specifically, the author aims to improve the existing error bound estimates to make them more accurate and effective.
### Research Background
The linear complementarity problem \( \text{LCP}(M, q) \) refers to finding a vector \( x \in \mathbb{R}^n \) such that:
\[
x \geq 0, \quad Mx + q \geq 0, \quad (Mx + q)^T x = 0
\]
where \( M = [m_{ij}] \in \mathbb{R}^{n \times n} \) and \( q \in \mathbb{R}^n \). This problem has wide applications in fields such as Nash equilibrium points, contact problems, and free - boundary problems.
It is known that when \( M \) is a P - matrix, \( \text{LCP}(M, q) \) has a unique solution for any \( q \in \mathbb{R}^n \). For B - matrices (a special type of P - matrices), there are already several error bound estimation methods in the literature, but these estimates may not be accurate enough or may be too conservative in some cases.
### Main Contributions of the Paper
This paper proposes a new error bound estimation method, which is applicable to the linear complementarity problem of B - matrices. The new method makes the error bound estimation more accurate by introducing more refined mathematical derivations and optimization techniques. Specifically, the paper proposes the following new error bound formula:
\[
\max_{d \in [0,1]^n} \|(I - D + DM)^{-1}\|_\infty \leq \sum_{i = 1}^n \frac{n - 1}{\min\{\bar{\beta}_i, 1\}} \prod_{j = 1}^{i - 1} \frac{b_{jj}}{\bar{\beta}_j}
\]
where \( \bar{\beta}_i \) is defined as follows:
\[
\bar{\beta}_i = b_{ii} - \sum_{j = i + 1}^n |b_{ij}| l_i(B^+), \quad l_k(B^+) = \max_{k \leq i \leq n} \left\{ \frac{1}{|b_{ii}|} \sum_{j = k, j \neq i}^n |b_{ij}| \right\}
\]
### Experimental Verification
The paper verifies the effectiveness of the new method through specific numerical examples. For example, for a given family of B - matrices \( M_k \), the error bound obtained by the new method is significantly better than the previous methods, especially when \( k \) is large, the error bound estimation is more accurate.
### Conclusion
The new error bound estimation method proposed in this paper significantly improves the existing results, especially when dealing with weakly diagonally dominant B - matrices. This provides more effective tools and methods for the research of linear complementarity problems.
In conclusion, the core problem of this paper is to improve the accuracy and reliability of the solutions to linear complementarity problems by improving the error bound estimates of B - matrices.