流形学习中的算法研究  

Research of Manifold Learning Algorithm

作　　者：周红[1] 吴炜[1] 滕奇志[1] 杨晓敏[1] 李旻[1] 陶德元[1] 

机构地区：[1]四川大学电子信息学院图像信息研究所,四川成都610064

出　　处：《计算机应用研究》2007年第7期214-217,共4页Application Research of Computers

摘　　要：详细介绍了一种新的机器学习的方法——流形学习。流形学习是一种新的非监督学习方法,可以有效地发现高维非线性数据集的内在维数并进行维数约简,近年来越来越受到机器学习和认知科学领域的研究者的重视。目前已经出现了很多有效的流形学习算法,如等度规映射(ISOMAP)、局部线性嵌套(Locally Linear Embedding,LLE)等。详细讲述了当前常用的几种流形学习算法以及在流形方面已经取得的研究成果,并对流形学习目前在各方面的应用作了较为细致的阐述。最后展望了流形学习的研究发展趋势,且提出了流形学习中仍需解决的关键问题。In the development of machine learning that to discover and study the orderliness from the data is always the main object of the machine learning. This paper illustrated the study achievement gained through a new method of machine learning- manifold learning in details. As a new unsupervised learning method ,manifold learning has captured the attention of many researchers in the field of machine learning and cognitive science. Currently, the major algorithms include isometric mapping （ISOMAP） and Locally Linear Embedding （LLE） , and so on. The approaches can be used for discovering the intrinsic dimensions of nonlinear high-dimensional data effectively and aim researchers to analyze the data better. It also gave a prospect of the development trend of the manifold learning research and proposed the crucial problems which need to be resolved in further work in manifold learning.

关 键 词：流形学习 主流形 局部线性嵌套 等度规映射 变分法 互信息 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]