A dual distance metrics method for improving classification performance

Abstract Distance metric forms the basis of pattern classification, as almost all classifiers depend on such a metric for making classification decisions. However, the existing research testifies that a single‐distance metric is not robust enough for classification. In this Letter, the authors propo...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Lian Wu, Yong Xu, Yong Zhao, Zhijun Hu, Lilei Sun
Formato:	article
Lenguaje:	EN
Publicado:	Wiley 2021
Materias:	Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/1743ddb7fea14b2bbd68bad9ad700cc5
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:1743ddb7fea14b2bbd68bad9ad700cc5
record_format	dspace
spelling	oai:doaj.org-article:1743ddb7fea14b2bbd68bad9ad700cc52021-11-16T10:18:22ZA dual distance metrics method for improving classification performance1350-911X0013-519410.1049/ell2.12016https://doaj.org/article/1743ddb7fea14b2bbd68bad9ad700cc52021-01-01T00:00:00Zhttps://doi.org/10.1049/ell2.12016https://doaj.org/toc/0013-5194https://doaj.org/toc/1350-911XAbstract Distance metric forms the basis of pattern classification, as almost all classifiers depend on such a metric for making classification decisions. However, the existing research testifies that a single‐distance metric is not robust enough for classification. In this Letter, the authors propose a dual distance metrics method and modify collaborative representation using dual distance metrics, that is, collaborative representation related class distance and conventional sample distance. These two distance metrics are fused by a parameter‐free multiplication scheme. The rationale of the designed multiplication fusion can be interpreted from the viewpoint of probability. As the multiplication fusion exploit the two distance metrics effectively and allow the best of the distance metrics to dominate the final classification decision, the improved collaborative representation based on dual distance metric can achieve a very high accuracy rate. Experimental results show that the proposed method has promising performance. It is also noted that the values of the two distance metrics vary with the test sample, therefore, the two distance metrics play different roles for different test samples. In other words, the two distance metrics are adaptive for test samples. The idea of the dual distance metrics method is also suitable for other combinations of other distance metrics.Lian WuYong XuYong ZhaoZhijun HuLilei SunWileyarticleElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENElectronics Letters, Vol 57, Iss 1, Pp 13-16 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Electrical engineering. Electronics. Nuclear engineering TK1-9971 Lian Wu Yong Xu Yong Zhao Zhijun Hu Lilei Sun A dual distance metrics method for improving classification performance
description	Abstract Distance metric forms the basis of pattern classification, as almost all classifiers depend on such a metric for making classification decisions. However, the existing research testifies that a single‐distance metric is not robust enough for classification. In this Letter, the authors propose a dual distance metrics method and modify collaborative representation using dual distance metrics, that is, collaborative representation related class distance and conventional sample distance. These two distance metrics are fused by a parameter‐free multiplication scheme. The rationale of the designed multiplication fusion can be interpreted from the viewpoint of probability. As the multiplication fusion exploit the two distance metrics effectively and allow the best of the distance metrics to dominate the final classification decision, the improved collaborative representation based on dual distance metric can achieve a very high accuracy rate. Experimental results show that the proposed method has promising performance. It is also noted that the values of the two distance metrics vary with the test sample, therefore, the two distance metrics play different roles for different test samples. In other words, the two distance metrics are adaptive for test samples. The idea of the dual distance metrics method is also suitable for other combinations of other distance metrics.
format	article
author	Lian Wu Yong Xu Yong Zhao Zhijun Hu Lilei Sun
author_facet	Lian Wu Yong Xu Yong Zhao Zhijun Hu Lilei Sun
author_sort	Lian Wu
title	A dual distance metrics method for improving classification performance
title_short	A dual distance metrics method for improving classification performance
title_full	A dual distance metrics method for improving classification performance
title_fullStr	A dual distance metrics method for improving classification performance
title_full_unstemmed	A dual distance metrics method for improving classification performance
title_sort	dual distance metrics method for improving classification performance
publisher	Wiley
publishDate	2021
url	https://doaj.org/article/1743ddb7fea14b2bbd68bad9ad700cc5
work_keys_str_mv	AT lianwu adualdistancemetricsmethodforimprovingclassificationperformance AT yongxu adualdistancemetricsmethodforimprovingclassificationperformance AT yongzhao adualdistancemetricsmethodforimprovingclassificationperformance AT zhijunhu adualdistancemetricsmethodforimprovingclassificationperformance AT lileisun adualdistancemetricsmethodforimprovingclassificationperformance AT lianwu dualdistancemetricsmethodforimprovingclassificationperformance AT yongxu dualdistancemetricsmethodforimprovingclassificationperformance AT yongzhao dualdistancemetricsmethodforimprovingclassificationperformance AT zhijunhu dualdistancemetricsmethodforimprovingclassificationperformance AT lileisun dualdistancemetricsmethodforimprovingclassificationperformance
_version_	1718426529617149952

A dual distance metrics method for improving classification performance

Ejemplares similares