计算机仿真
計算機倣真
계산궤방진
COMPUTER SIMULATION
2010年
2期
145-148,268
,共5页
网格技术%关联规则挖掘%并行挖掘特性
網格技術%關聯規則挖掘%併行挖掘特性
망격기술%관련규칙알굴%병행알굴특성
Grid technologies%Association rules mining%Parallel mining characteristic
科学和工商业应用需要分析分布在各异构站点的海量数据.传统的关联规则挖掘算法探讨的对象基本上都是集中式的数据集,对分布式的动态数据库群无能为力,因而迫切需要对分布式数据挖掘算法进行研究探讨.在研究OGSA面向服务的体系结构基础上,将网格技术与数据挖掘技术有机地结合在一起,提出了一种基于网格的分布式关联规则挖掘方法.是改进Apriori挖掘算法在网格环境下的具体应用.仿真实验表明方法具备网格的并行挖掘特性,能够成功实现位于多个异构站点上的分布式数据挖掘,且挖掘速度和运算效率较之集中式Apriori挖掘算法有较大幅度的提高.
科學和工商業應用需要分析分佈在各異構站點的海量數據.傳統的關聯規則挖掘算法探討的對象基本上都是集中式的數據集,對分佈式的動態數據庫群無能為力,因而迫切需要對分佈式數據挖掘算法進行研究探討.在研究OGSA麵嚮服務的體繫結構基礎上,將網格技術與數據挖掘技術有機地結閤在一起,提齣瞭一種基于網格的分佈式關聯規則挖掘方法.是改進Apriori挖掘算法在網格環境下的具體應用.倣真實驗錶明方法具備網格的併行挖掘特性,能夠成功實現位于多箇異構站點上的分佈式數據挖掘,且挖掘速度和運算效率較之集中式Apriori挖掘算法有較大幅度的提高.
과학화공상업응용수요분석분포재각이구참점적해량수거.전통적관련규칙알굴산법탐토적대상기본상도시집중식적수거집,대분포식적동태수거고군무능위력,인이박절수요대분포식수거알굴산법진행연구탐토.재연구OGSA면향복무적체계결구기출상,장망격기술여수거알굴기술유궤지결합재일기,제출료일충기우망격적분포식관련규칙알굴방법.시개진Apriori알굴산법재망격배경하적구체응용.방진실험표명방법구비망격적병행알굴특성,능구성공실현위우다개이구참점상적분포식수거알굴,차알굴속도화운산효솔교지집중식Apriori알굴산법유교대폭도적제고.
Scientific and industrial applications require analysis of the distribution sites in the heterogeneous mass of data.The traditional association rule mining algorithms to explore the targets are basically centralized data sets,and can do nothing to the group's distributed dynamic database.So it is urgent to carry out research to explore distributed data mining algorithms.After the service-oriented architecture of OGSA is studied,the paper combines the Grid and data mining technology organically,and then proposes an association rules mining method based on Grid.The method is an idiographic application of the improved Apriori mining algrithom in the gridcircumstance.The simulation shows that the method meshes with the parallel mining characteristic,achieves success at the site of a number of heterogeneous distributed data mining,and makes a sharp increase in speed and efficiency of mining and computing compared with centralized Apriori mining algorithm.