节点文献
多概念层次上基于赋范划分距离的分类算法
Classify Algorithm Based on Normal Partition Distance at Multiple Concept Levels
【摘要】 应用知识发现理论与方法从大型考生信息数据库中提取隐含的、前所未知的、对决策有潜在价值的知识和规则,是当前计算机教育 应用的新课题。根据网上远程考试与评价系统的特点,讨论了挖掘的目标规则类型及其挖掘算法。针对算法的不足,提出了一种新的分ID3类算法多概念层次上基于赋范划分距离的分类算法,简要介绍了在知识发现过程中的应用。--XML
【Abstract】 Application of KDD technology to extract implicit,previously unknown,and potentially useful rules from very large databases withexaminees’ information is a new task in the computer eductional application category.Based on the characteristics of distance test and evaluation system,we first discuss the result rules’ kinds and relative mining algorithms.Then a new classify algorithm based on normal partition distance at multiple concept levels is addressed to overcome ID3’s shortcoming.Finally application of XML in the complete process of KDD is briefly introduced.
【Key words】 KDD; Data mining; XML; Induction learning; Normal partition distance; Classify algorithm;
- 【文献出处】 计算机工程 ,Computer Engineering , 编辑部邮箱 ,2001年07期
- 【分类号】G434
- 【被引频次】2
- 【下载频次】103