文献类型：期刊文献

中文题名：基于随机游走的流形学习与可视化

英文题名：Manifold Learning and Visualization Based on Random Walk

作者：邵超[1];万春红[1];张啸剑[1]

第一作者：邵超

机构：[1]河南财经政法大学计算机与信息工程学院

第一机构：河南财经政法大学计算机与信息工程学院

年份：2017

卷号：32

期号：3

起止页码：559-569

中文期刊名：数据采集与处理

外文期刊名：Journal of Data Acquisition and Processing

收录：CSTPCD;;Scopus;北大核心:【北大核心2014】；CSCD:【CSCD_E2017_2018】；

基金：国家自然科学基金(61202285)资助项目;河南省教育厅科学技术研究重点(14B520020)资助项目

语种：中文

中文关键词：全局流形学习;等距映射;邻域图;随机游走;通勤时间距离

外文关键词：global manifold learning; isometric mapping; neighborhood graph; random walk; commutetime distance

摘要：现有的全局流形学习算法都敏感于邻域大小这一难以高效选取的参数,它们都采用了基于欧氏距离的邻域图创建方法,从而使邻域图容易产生"短路"边。本文提出了一种基于随机游走模型的全局流形学习算法(Random walk-based isometric mapping,RW-ISOMAP)。和欧氏距离相比,由随机游走模型得到的通勤时间距离是由给定两点间的所有通路以概率为权组合而成的,不但鲁棒性更高,而且还能在一定程度上度量具有非线性几何结构的数据之间的相似性。因此采用通勤时间距离来创建邻域图的RW-ISOMAP算法将不再敏感于邻域大小参数,从而可以更容易地选取邻域大小参数,同时还具有更高的鲁棒性。最后的实验结果证实了该算法的有效性。
The existing global manifold learning algorithms are relatively sensitive to the neighborhood size, which is difficult to select efficiently. The reason is mainly because the neighborhood graph is con- structed based on Euclidean distance, by which shortcut edges tend to be introduced into the neighbor- hood graph. To overcome this problem, a global manifold learning algorithm is proposed based on ran- dom walk, called the random walk-based isometric mapping （RW-ISOMAP）. Compared with Euclidean distance, the commute time distance, achieved by the random walk on the neighborhood graph, can measure the similarity between the given data within the nonlinear geometric structure to a certain ex- tent, thus it can provide robust results and is more suitable to construct the neighborhood graph. Conse- quently, by constructing the neighborhood graph based on the commute time distance, RW-ISOMAP is less sensitive to the neighborhood size and more robust than the existing global manifold learning algo- rithms. Finally, the experiment verifies the effectiveness of RW-ISOMAP.