摘要
介绍了AprioriHybird算法,用2项集支持度矩阵对其生成频繁2项集和频繁3项集的方法进行了改进,同时对数据库消减也给予了一定的策略,最后通过股市数据的实验,证明了改进算法的效率在一定程度上优于AprioriHybird算法;同时挖掘出大量有意义的关联规则,用于指导模拟买卖取得了较佳效果。
The paper introduces AprioriHybird algorithm, and improves the method of making 2-frequent itemsets and 3-frequent itemsets with 2-itemsets support matrix, and gives some strategies on reducing database. Finally, the experiments of stock market data show that the efficiency of the improved AprioriHybird was superior to AprioriHybird algorithm to a certain extent. A lot of significant association rules were extracted. Good effect was obtained in instructing simulated trade by them.
出处
《电脑开发与应用》
2007年第10期36-38,共3页
Computer Development & Applications
基金
四川省重大基础研究项目子课题基金(04JY029-001-4)