| 摘要: |
| 对关联规则的增量式更新算法——IUA(incremental updating algorithm)进行了分析,指出其存在的问题,并给出一个改进的算法NEWIUA(new IUA).NEWIUA算法对已存在的和本次更新时新产生的频繁项目集都加以充分的利用,因此,在保证算法有效的同时提高了效率.文章提出了3种关联规则开采的并行算法,并对各算法进行了分析和讨论. |
| 关键词: 数据开采,关联规则,增量式更新,频繁项目集,并行算法. |
| DOI： |
| 分类号: |
| 基金项目:本文研究得到国家自然科学基金和太原师范专科学校科研基基金资助。 |
|
| Data Mining and Incremental Updating on Association Rules |
|
ZHOU Hai-yan
|
| Abstract: |
| In this paper, the author analyzes the incremental updating algorithm (IUA) on association rules, points out its existing problems, and presents an improved algorithm, NEWIUA (new IUA), which takes full use of already existing and the current updated new frequent itemsets, therefore the efficiency is increased besides guaranteeing the validity of the algorithm. Three parallel algorithms for data mining on association rules are presented, the analysis and discussion on each algorithm are also presented. |
| Key words: Data mining, association rules, incremental updating, frequent itemsets, parallel algorithm. |
