An Evolutionary Associative Contrast Rule Mining Method for Incomplete Database
An Evolutionary Associative Contrast Rule Mining Method for Incomplete Database
复制标题
DOI：
--
复制
发表时间：
2013
期刊：
影响因子：
--
通讯作者：
K. Shimada;T. Hanioka
中科院分区：
文献类型：
--
作者：
K. Shimada;T. Hanioka
提出了一种从不完备数据库中挖掘关联对比规则的方法，以发现两个不完整数据集之间有趣的差异。该方法提取类似“If X Then Y”的规则，这些规则只在聚焦类中才有意义。该方法采用了基于进化图优化技术的基本结构，并采用了一种新的进化策略来在其进化过程中积累规则。该方法可以利用卡方检验实现两类不完备数据库之间的关联分析。我们评估了进化方法在不完备数据库中挖掘关联对比规则的性能。此外，还对缺失值对规则测量的危害进行了评价。
A method for associative contrast rule mining from incomplete database is demonstrated to find interesting differences between two incomplete data sets. The method extracts rules like "if X then Y" is interesting only in the focusing class. The method has been developed using a basic structure of the evolutionary graph-based optimization technique and adopting a new evolutionary strategy to accumulate rules through its evolutionary process. The method can realize the association analysis between two classes of the incomplete database using chi-square test. We evaluated the performance of the evolutionary method for associative contrast rule mining for the incomplete database. In addition, the evaluation of the mischief for the rule measurements by missing values is demonstrated.