Journal Published Online: 21 November 2017
Volume 46, Issue 3

Principal Component Analysis Based on Marginal Density Ratios

CODEN: JTEVAB

Abstract

Principal component analysis is a classical technique for feature extraction and dimension reduction. But it only exploits the input data and does not consider other assisting information. In the paper, we propose a novel principal component analysis based on marginal density ratios. Motivating by a naïve Bayes model, the log ratio of joint densities for features of samples is incorporated to the transform sample matrix. The acquired sample matrix is then used to calculate principal components using the singular value decomposition. Finally, the k-nearest neighbor is used as the classified method. Experimental results on synthetic and real data sets demonstrate promising performance both in accuracy and computational efficiency.

Author Information

Shi, W.
College of Informational Science and Engineering, Henan University of Technology, Henan, China Collaborative Innovation Center of Grain Storage Security, Henan University of Technology, Henan, China
Jiang, J.
Department of Mathematics and Statistics, University of North Carolina at Charlotte, NC
Pages: 7
Price: $25.00
Related
Reprints and Permissions
Reprints and copyright permissions can be requested through the
Copyright Clearance Center
Details
Stock #: JTE20160094
ISSN: 0090-3973
DOI: 10.1520/JTE20160094