一种新的不确定树模式聚类算法

摘要

不确定树模式聚类是数据挖掘领域中的一个重要问题,提出了一种新的不确定树模式聚类算法,有效地解决了因数据的不确定性而导致的无法聚类的问题.为了更加准确地度量树模式之间的相似性,提出了一种语义相似度计算方法与结构相似度计算方法.设计了一个动态聚类过程,自适应获取聚类阈值,较大程度上减少了人为干扰导致聚类结果不准确的影响,使得具有相似结构的子树聚集在同一个相似分组中,不同分组之间的子树相似度达到最小化.通过模拟数据和真实环境两部分实验表明,算法有效可行,聚类结果较准确且具有较好的运行效率.%Uncertain tree clustering is an important problem in data mining domain.In this paper,a new uncertain tree clustering algorithm is proposed.The algorithm effectively resolves the clustering problems for uncertain data.In order to improve accurate measurement on the similarities among trees,the method of semantic similarity and structural similarity are presented.A dynamic clustering process is designed in which self-adaptive threshold be applied so as to greatly reduce the jamming impact on the result accuracy.This process can cluster subtrees of similar structure within similar groups,minimizing the similarity of subtree groups.Both simulation and real experiments show that the algorithm is effective and efficient and the clustering result is accurate.