View in EDS

Constrained Laplacian Eigenmap for dimensionality reduction

Saved in:
Bibliographic Details
Title: Constrained Laplacian Eigenmap for dimensionality reduction
Authors: Chen, Chun1 chenc@zju.edu.cn, Zhang, Lijun1 zljzju@zju.edu.cn, Bu, Jiajun1 bjj@zju.edu.cn, Wang, Can wcan@zju.edu.cn, Chen, Wei1 chenw@zju.edu.cn
Source: Neurocomputing. Jan2010, Vol. 73 Issue 4-6, p951-958. 8p.
Subjects: MAP (Computer program language), Dimension reduction (Statistics), Laplacian operator, Graphic methods, Embeddings (Mathematics), Algorithms, Constraint satisfaction
Abstract: Abstract: Dimensionality reduction is a commonly used tool in machine learning, especially when dealing with high dimensional data. We consider semi-supervised graph based dimensionality reduction in this paper, and a novel dimensionality reduction algorithm called constrained Laplacian Eigenmap (CLE) is proposed. Suppose the data set contains r classes, and for each class we have some labeled points. CLE maps each data point into r different lines, and each map i tries to separate points belonging to class i from others by using label information. CLE constrains the solution space of Laplacian Eigenmap only to contain embedding results that are consistent with the labels. Then, each point is represented as a r-dimensional vector. Labeled points belonging to the same class are merged together, labeled points belonging to different classes are separated, and similar points are close to one another. We perform semi-supervised document clustering using CLE on two standard corpora. Experimental results show that CLE is very effective. [Copyright &y& Elsevier]
Copyright of Neurocomputing is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:Abstract: Dimensionality reduction is a commonly used tool in machine learning, especially when dealing with high dimensional data. We consider semi-supervised graph based dimensionality reduction in this paper, and a novel dimensionality reduction algorithm called constrained Laplacian Eigenmap (CLE) is proposed. Suppose the data set contains r classes, and for each class we have some labeled points. CLE maps each data point into r different lines, and each map i tries to separate points belonging to class i from others by using label information. CLE constrains the solution space of Laplacian Eigenmap only to contain embedding results that are consistent with the labels. Then, each point is represented as a r-dimensional vector. Labeled points belonging to the same class are merged together, labeled points belonging to different classes are separated, and similar points are close to one another. We perform semi-supervised document clustering using CLE on two standard corpora. Experimental results show that CLE is very effective. [Copyright &y& Elsevier]
ISSN:09252312
DOI:10.1016/j.neucom.2009.08.021