面向分类预测的增量关联规则更新研究
Research on Incremental Updating Algorithms for Mining Classification Association Rules
-
摘要: 在对现有增量关联规则更新算法进行讨论的基础上, 针对只关注分类预测结果中的某些特定类别的应用, 提出了一个改进算法.该算法通过对类别结果的限制, 有效减少了算法在多维增量关联规则挖掘过程中扫描数据库的次数及使用的候选谓词集表中记录的个数, 使算法具有更优的空间复杂度、更高的结果聚焦度.Abstract: In some applications of classification, the users just give attention to one of the results.For these applications, it improved the "Practically high efficiency calculating method in association rules and data mining".Through the improvement, the numbers of itemsets are much less than before.
下载:
