Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/31302
Title: On kernel difference-weighted k-nearest neighbor classification
Authors: Zuo, W
Zhang, D 
Wang, K
Keywords: Classifier
Distance-weighted KNN
Kernel methods
Nearest neighbor
Pattern classification
Issue Date: 2008
Source: Pattern analysis and applications, 2008, v. 11, no. 3-4, p. 247-257 How to cite?
Journal: Pattern Analysis and Applications 
Abstract: Nearest neighbor (NN) rule is one of the simplest and the most important methods in pattern recognition. In this paper, we propose a kernel difference-weighted k-nearest neighbor (KDF-KNN) method for pattern classification. The proposed method defines the weighted KNN rule as a constrained optimization problem, and we then propose an efficient solution to compute the weights of different nearest neighbors. Unlike traditional distance-weighted KNN which assigns different weights to the nearest neighbors according to the distance to the unclassified sample, difference-weighted KNN weighs the nearest neighbors by using both the correlation of the differences between the unclassified sample and its nearest neighbors. To take into account the effective nonlinear structure information, we further extend difference-weighted KNN to its kernel version KDF-KNN. Our experimental results indicate that KDF-WKNN is much better than the original KNN and the distance-weighted KNN methods, and is comparable to or better than several state-of-the-art methods in terms of classification accuracy.
URI: http://hdl.handle.net/10397/31302
ISSN: 1433-7541
DOI: 10.1007/s10044-007-0100-z
Appears in Collections:Journal/Magazine Article

Access
View full-text via PolyU eLinks SFX Query
Show full item record

SCOPUSTM   
Citations

43
Last Week
0
Last month
0
Citations as of Sep 22, 2017

WEB OF SCIENCETM
Citations

36
Last Week
0
Last month
0
Citations as of Sep 22, 2017

Page view(s)

49
Last Week
4
Last month
Checked on Sep 25, 2017

Google ScholarTM

Check

Altmetric



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.