摘要
为了深入地探索聚类结果簇的形态特征,提出了一种基于维度映射的类圆簇识别算法。该算法将结果簇按维度进行映射,通过比较、分析簇在各个映射维度上的频数曲线及形态特征,自动将类圆簇从众多结构复杂的聚类结果簇中识别出来。算法经过大量实验验证,具有很好的识别能力和抗干扰能力,对于高维度数据集合也具有很强的扩展性。
In order to explore the real features of outcome clusters,this paper proposed a suborbicular cluster recognition algorithm based on multi-dimensional analysis.Firstly,it analyzed the given data sets in each dimension,then drew the frequency cure of every dimension and compared the characteristics of those cures.Finally,it identified the suborbicular cluster from a number of clusters.A large number of experiments show that the algorithm is robust,effective.It can be extended to high-dimensional data sets.
出处
《计算机应用研究》
CSCD
北大核心
2011年第10期3665-3670,3674,共7页
Application Research of Computers
关键词
聚类簇识别
维度分析
聚类分析
数据挖掘
cluster recognition
analysis of data dimension
cluster analysis
data mining
