一种基于信息粒度的高属性维稀疏数据聚类算法

目前的高属性维稀疏数据算法大多面向二态数据，而且没有聚类结果的评价方法，给应用带来很大局限。针对这些问题，给出一种基于信息粒度的高属性维聚类算法。首先通过设计面向数据稀疏特征的半模糊聚类算法对数据进行离散化，并基于此给出稀疏相似度和初始等价关系的定义，然后设计可变精度的二次聚类模型对初始聚类结果进行修正，使算法具有较强抗噪声能力，最后结合应用领域定义一种新的聚类质量的评价模型。实验证明，算法具有更广应用性，可提供多粒度分析结果，准确度更高，得到的聚类结果能真实反映数据的特征。

A Knowledge granular based High Attribute Dimensional Sparse Clustering

Currently a majority of high attribute dimensional sparse clustering algorithms can only handle binary data and lack of evaluation method for clustering results, which brings great limits to applications. To solve these problems, this paper brings forward a knowledge granular based clustering algorithm. Through a semi-fuzzy algorithm the sparse data are discretize. Based on these, sparse similarity and initial equivalence relation are defined and Variable Precision quadratic clustering model is designed to refine the result so the algorithm gains noise resistance ability. A new clustering quantity evaluation model is defined facing the application field. The test has shown that the algorithm provides results of various granular with high veracity and shows the data characteristics.