New Methods for Spectral Clustering.

New Methods for Spectral Clustering.
复制标题

谱聚类的新方法。

DOI:
--
复制
发表时间:
2004
期刊:
影响因子:
--
通讯作者:
J. Poland
J. Poland
中科院分区:
--
文献类型:
--
作者:
Igor Fischer;J. Poland

文献摘要

被引文献

相似文献

分析亲和矩阵谱是一种日益流行的数据聚类方法。我们提出了三个新的算法组件,这是适当的,以提高性能的谱聚类。首先,观察特征向量建议使用K线算法而不是常用的K均值。其次,如果亲和矩阵具有清晰的块结构,则聚类效果最好,这可以通过计算电导率矩阵来实现。第三,许多聚类问题是不均匀的或不对称的,在这个意义上,一些集群是集中的,而另一些是分散的。在这种情况下,关联矩阵的上下文相关计算会有所帮助。该方法还证明允许鲁棒的核半径σ的自动确定。
Analyzing the affinity matrix spectrum is an increasingly popular data clustering method. We propose three new algorithmic components which are appropriate for improving performance of spectral clustering. First, observing the eigenvectors suggests to use a K-lines algorithm instead of the commonly applied K-means. Second, the clustering works best if the affinity matrix has a clear block structure, which can be achieved by computing a conductivity matrix. Third, many clustering problems are inhomogeneous or asymmetric in the sense that some clusters are concentrated while others are dispersed. In this case, a context-dependent calculation of the affinity matrix helps. This method also turns out to allow a robust automatic determination of the kernel radius σ.