Diffusion maps

Diffusion maps
复制标题

DOI:
10.1016/j.acha.2006.04.006
复制
发表时间:
2006-07-01
影响因子:
2.5
通讯作者:
Lafon, Stephane
Lafon, Stephane
中科院分区:
数学1区
文献类型:
--
作者:
Coifman, Ronald R.;Lafon, Stephane

文献摘要

被引文献

相似文献

在本文中,我们提供了一个基于扩散过程的框架,用于寻找数据集的有意义的几何描述。我们证明,马尔可夫矩阵的特征函数可用于构造称为扩散图的坐标,从而生成复杂几何结构的有效表示。通过迭代马尔可夫矩阵获得的相关扩散距离族定义了多尺度几何形状,这些几何形状被证明在数据参数化和降维方面很有用。所提出的框架将马尔可夫过程的谱特性与其几何对应物联系起来,并且统一了机器学习、谱图理论和特征图方法等各种背景下产生的想法。 (C) 2006 年,爱思唯尔公司出版。
In this paper, we provide a framework based upon diffusion processes for finding meaningful geometric descriptions of data sets. We show that eigenfunctions of Markov matrices can be used to construct coordinates called diffusion maps that generate efficient representations of complex geometric structures. The associated family of diffusion distances, obtained by iterating the Markov matrix, defines multiscale geometries that prove to be useful in the context of data parametrization and dimensionality reduction. The proposed framework relates the spectral properties of Markov processes to their geometric counterparts and it unifies ideas arising in a variety of contexts such as machine learning, spectral graph theory and eigenmap methods. (C) 2006 Published by Elsevier Inc.