摘要
目的求解n维空间中m个点的最小闭包球(MEB)问题。方法基于序列最小优化(SMO)的方法,提出了一种近似算法,求解MEB问题的一个(1+ε)-近似。结果建立了此算法的计算复杂度为O(mn/ε),并且算法最终得到一个独立于m,n的大小为O(1/ε)的核心集。结论数值结果表明对于求解高精度的大规模问题,算法是很有效的。
Aim To solve the minimum enclosing ball(MEB) problem of m points in n dimensions.Methods Based on the approach of sequential minimal optimization(SMO),an approximation algorithm for the MEB problem is proposed,in order to get a(1+ε)-approximation to the MEB problem.Results This algorithm computes a(1+ε)-approximation to MEB in O(mn/ε) arithmetic operations and achieves a core set of size O(1/ε),which is independent of both m and n.Conclusion The numerical results demonstrate the efficiency of the method for solving large-scale problem with a high degree of accuracy.
出处
《西北大学学报(自然科学版)》
CAS
CSCD
北大核心
2010年第6期965-969,共5页
Journal of Northwest University(Natural Science Edition)
基金
国家自然科学基金资助项目(61072144)
中央高校基本科研业务费基金资助项目(JY10000970004)
关键词
最小闭包球
序列最小优化
近似算法
计算复杂度
核心集
minimum enclosing balls
sequential minimal optimization
approximation algorithms
computational complexity
core sets