基于矩阵加权关联规则的核模糊聚类

Fuzzy Core Clustering Based on Matrix of Weighted Asociation Rules

提出一种基于矩阵加权关联规则的空间粒度聚类算法。该算法核心思想是根据文档特征向量矩阵提取文档的相似度,再在该关联规则算法上进行聚类来寻找相似关系的频繁项集。通过引入核函数,样本点被非线性变换映射到高维特征空间进行聚类,提高聚类性能。通过矩阵加权关联规则算法进行聚类。通过实验表明,在处理中小型文档时,该算法的精确度优于传统Apriori算法和K-mean算法;在处理大型文档时,该算法的时间复杂度小于传统的K-mean算法。

This paper presents a spatial granularity clustering algorithm based on matrix of weighted association rules.The core idea of the algorithm is based on the document feature vectors matrix extracted document similarity,then by the association rules algorithms for clustering to find similar relations between the frequent itemsets.By introducing the kernel function,the sample points mapped by non-linear transformation into high-dimensional feature space for clustering,and clustering performance been improved.Matrix of weighted association rules by clustering algorithms.Experiments show that,in dealing with small and medium sized document,the accuracy of the algorithm is superior to the traditional Apriori algorithm and K-mean algorithm;in dealing with large documents,the algorithm＇s time complexity is less than the traditional K-mean algorithm.