基于树的不确定数据频繁项的挖掘算法研究

Researches on Tree-based Frequent Pattern Mining from Uncertain Data

本文针对UF-growth算法在构建环节的不足,提出了一种改进的映射频繁项树PUF-tree算法。该算法基于数据库映射技术,通过改进原算法的构建过程,能在一定程度上消除原算法的性能瓶颈,提高算法的效率。

According to the deficiency of the traditional UF-growth, an improved algorithm is introduced into the Projection Frequent Pattern Tree （PUF-tree） from Uncertain data. The PUF-tree method, based on the database projection technology, can avoid the bottleneck of the conventional algorithm and works more efficiently by using an improved tree constructing process.