一种基于预裁剪的不确定数据流频繁项集挖掘算法

随着交通、网络流量监控等应用的涌现,不确定数据流频繁项集挖掘成为近年来的研究热点。通常在不确定数据流中,频繁项集所占的比例较小,导致挖掘中无效操作较多。基于这种情况,提出了一种基于预裁剪的不确定数据流频繁项集挖掘算法Prep-UF-Streaming;该算法,不仅能裁剪掉大部分非频繁项集,提高了算法的平均运行时间;而且能够检测到非频繁项集成为频繁项集的可能性,尽量不丢失频繁项集,从而尽可能地提高算法的性能。

An Algorithm of Frequent Itemset Mining over Uncertain Data Streams based on Pre-pruning

With the emergence of applications such as traffic, network traffic monitoring, frequent itemset mining over uncertain data streams becomes hotspot in recent years. Usually the percentage of frequent itemset in the uncertain data streams is relatively small. Based on this, we proposed an algorithm of frequent itemset mining over uncertain data streams based on pre-pruning named Prep-UF-Streaming. This algorithm not only prunes most of the non-frequent itemset and increases the average running time, but also can detect the possibility of non-frequent become frequent itemset, and try not to lose frequent itemset, thus as far as possible to improve its efficiency.