一类随机线性互补问题的求法 |
| |
引用本文: | 薛文娟,朱彬,钟一文. 一类随机线性互补问题的求法[J]. 数学的实践与认识, 2014, 0(5) |
| |
作者姓名: | 薛文娟 朱彬 钟一文 |
| |
作者单位: | 福建农林大学计算机与信息学院;福建师范大学经济学院; |
| |
基金项目: | 福建农林大学校青年教师基金(2011xjj24);福建省教育厅科技项目(JB12066);国家社科基金项目(12CTQ031);教育部人文社科规划基金项目(10YJCZH249);福建省出国留学奖学金资助 |
| |
摘 要: | ![]() 考虑一类随机互线性补问题的求解方法,目的是通过定义NCP函数来使正则化期望残差最小化.通过拟蒙洛包洛方法产生一系列观察值并且证得离散近似问题最小值解的聚点就是相应随机线性互补问题的期望残差最小值ERM,同时得到利用ERM到解为有界的充分条件.进一步证明ERM法能够得到具有稳定性和最小灵敏度的稳健解.
|
关 键 词: | 随机线性互补问题 正则化期望残差最小化 蒙洛包洛方法 |
The Solution of Stochastic Linear Complementarity Problems |
| |
Abstract: | ![]() The paper presents a formulation for the stochastic linear complementarity problem,which aims at minimizing a regularized expected residual by defining an NCP function.It generates observations by the quasi-Monte Carlo methods and proves that every accumulation point of minimizer of discrete approximation problems is a minimum expected residual minimization(ERM) problem.Furthermore,it gets a sufficient condition for the existence of bounded of solution which is solved by the regularized ERM method.Furthermore,the ERM formulation produces robust solutions Which minimum,high reliability. |
| |
Keywords: | stochastic linear complementary problem regularized expected residual minimization Monte Carlo method |
本文献已被 CNKI 等数据库收录! |