两种流形学习算法的对比研究  

A comparative study of two kinds of manifold learning algorithms

作　　者：王博[1] 刘美玲[1] 张学敏[1] 

机构地区：[1]西安建筑科技大学,陕西西安710055

出　　处：《微型机与应用》2013年第8期42-44,共3页Microcomputer & Its Applications

摘　　要：介绍了局部线性嵌套和等距映射两种最基本的非线性降维方法,对比测试了两种降维方法在不同参数下的执行效果与效率,总结了两种降维方法所适合的数据特点,并应用于图像识别中,比较了两者在图像识别中的识别率。Data dimensionality reduction ingeneral can be divided into linear dimensionality reduction and nonlinear dimensional- ity reduction,locally linear embedding and isometric mapping is one of the two most recently the nonlinear dimensionality reduction method,for different sets of data dimensionality reduction have their advantages and disadvantages.The contrast test of weo dimension reduction method under different parameters of effectiveness and efficiency,summarizes itsdimensionality reduction fit for the charac- teristics of data, is aoofied to image recognition,co,npared to the respective image recognition in recognition rate.

关 键 词：非线性降维 流形学习 局部线性嵌套 等距映射 人脸识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]