一种多段支持度数据挖掘算法

A Data Mining Algorithm With Multi-segments Support

本文以分段计算支持度为重要思想,通过分段计算各项集支持度,确保各段记录出现在相应规模事务中所形成的频度,进而构成支持度向量,加上项集多段支持度,实现大规模频繁项集的有效推测。该算法可提高数据库扫描过程中的信息获取率,从而缩减项集规模,并按照文中定理1这一思想对数据集进行及时调整,从而实现频繁项集生成效率的不断提高。

Based on the thought of piecewise calculation support, this papercalculatesevery itemsets’ support degreeto find the frequency ofallsegments formedin correspondingtransaction, thusconstitutes thesupport vector. With this,plusmulti-segmentsupport of sets,effective speculation inlarge-scalefrequent item sets is achieved.The algorithmwillimprove database’saccess to informationduring the scanrate,thereby reducing the scale ofitemsets. If dataset is adjusted according to theory one in the paper,generationefficiency of frequent itemset will continue to increase.