A comparison of SOM neural network and hierarchical clustering methods
A comparison of SOM neural network and hierarchical clustering methods
复制标题
DOI：
10.1016/0377-2217(96)00038-0
复制
发表时间：
1996-09-06
影响因子：
6.4
通讯作者：
West, D
中科院分区：
文献类型：
--
作者：
Mangiameli, P;Chen, SK;West, D
聚类分析，确定数据集中的自然亚组，是一种重要的统计方法，在许多情况下使用。今天使用的层次聚类方法的一个主要问题是，当经验数据偏离紧凑孤立集群的理想条件时，会发生分类错误的趋势。许多经验数据集有结构上的缺陷，混淆了集群的识别。我们使用自组织映射（SOM）神经网络聚类方法，并证明它是上级的层次聚类方法。的神经网络和七个层次聚类方法的性能进行了测试，252个数据集，不同程度的不完善，包括数据分散，离群值，不相关的变量，和不均匀的集群密度。神经网络优越的上级精度和鲁棒性可以提高基于聚类杂乱经验数据的决策和研究的有效性。
Cluster analysis, the determination of natural subgroups in a data set, is an important statistical methodology that is used in many contexts. A major problem with hierarchical clustering methods used today is the tendency for classification errors to occur when the empirical data departs from the ideal conditions of compact isolated clusters. Many empirical data sets have structural imperfections that confound the identification of clusters. We use a Self Organizing Map (SOM) neural network clustering methodology and demonstrate that it is superior to the hierarchical clustering methods. The performance of the neural network and seven hierarchical clustering methods is tested on 252 data sets with various levels of imperfections that include data dispersion, outliers, irrelevant variables, and nonuniform cluster densities. The superior accuracy and robustness of the neural network can improve the effectiveness of decisions and research based on clustering messy empirical data.