Advanced Techniques in Computing Sciences and Software Engineering
pp 7984
Date:
Comparative Study of Distance Functions for Nearest Neighbors
 Janett WaltersWilliamsAffiliated withSchool of Computing and Information Technology, University of Technology Email author
 , Yan LiAffiliated withDepartment of Mathematics and Computing Centre for Systems Biology, University of Southern Queensland
Abstract
Many learning algorithms rely on distance metrics to receive their input data. Research has shown that these metrics can improve the performance of these algorithms. Over the years an often popular function is the Euclidean function. In this paper, we investigate a number of different metrics proposed by different communities, including Mahalanobis, Euclidean, KullbackLeibler and Hamming distance. Overall, the bestperforming method is the Mahalanobis distance metric.
Keywords
KullbackLeibler distance Euclidean distance Mahalanobis distance Manhattan distance Hamming distance Minkowski distance Nearest Neighbor Title
 Comparative Study of Distance Functions for Nearest Neighbors
 Book Title
 Advanced Techniques in Computing Sciences and Software Engineering
 Pages
 pp 7984
 Copyright
 2010
 DOI
 10.1007/9789048136605_14
 Print ISBN
 9789048136599
 Online ISBN
 9789048136605
 Publisher
 Springer Netherlands
 Copyright Holder
 Springer Science+Business Media B.V.
 Additional Links
 Topics
 Keywords

 KullbackLeibler distance
 Euclidean distance
 Mahalanobis distance
 Manhattan distance
 Hamming distance
 Minkowski distance
 Nearest Neighbor
 Industry Sectors
 eBook Packages
 Editors

 Khaled Elleithy ^{(ID1)}
 Editor Affiliations

 ID1. School of Engineering, University of Bridgeport
 Authors

 Janett WaltersWilliams ^{(1)}
 Yan Li ^{(2)}
 Author Affiliations

 1. School of Computing and Information Technology, University of Technology, Kingston 6, Jamaica W.I
 2. Department of Mathematics and Computing Centre for Systems Biology, University of Southern Queensland, Toowoomba, Australia
Continue reading...
To view the rest of this content please follow the download PDF link above.