一种基于证据距离的多分类器差异性度量

电子与自动控制

本期目录 | 过刊浏览 | 高级检索

前一篇 | 后一篇

一种基于证据距离的多分类器差异性度量

杨艺¹, 韩德强², 韩崇昭²

1.西安交通大学航天航空学院, 陕西西安 710049;
2. 西安交通大学电子与信息工程学院综合自动化研究所, 陕西西安 710049

收稿日期:2011-08-31 修回日期:2011-10-24 出版日期:2012-06-25 发布日期:2012-06-26
通讯作者: 韩德强 E-mail:deqhan@mail.xjtu.edu.cn
基金资助:
国家"973"计划(2007CB311006);国家自然科学基金(61104214, 61074176, 67114022);中国博士后科学基金(20100481337, 201104670);陕西省电子信息系统综合集成重点实验室基金(201101Y17);重庆市自然科学基金(CSCT, 2010BA2003)

A Novel Diversity Measure of Multiple Classifier Systems Based on Distance of Evidence

YANG Yi¹, HAN Deqiang², HAN Chongzhao²

1. School of Aerospace, Xi’an Jiaotong University, Xi’an 710049, China;
2. Institute of Integrated Automation, School of Electronic and Information Engineering, Xi’an Jiaotong University, Xi’an 710049, China

Received:2011-08-31 Revised:2011-10-24 Online:2012-06-25 Published:2012-06-26
Supported by:
National Basic Research Program of China (2007CB311006); National Natural Science Foundation of China (61104214, 61074176, 67114022); China Postdoctoral Science Foundation (20100481337, 201104670); Research Fund of Shanxi Key Laboratory of Electronic Information System Integration (201101Y17); Chongqing Natural Science Foundation(CSCT, 2010BA2003)

摘要/Abstract

摘要： 多分类器系统因其能够显著提升分类精度而引发了广泛关注。多分类器系统中各子分类器间的差异性是提升融合分类精度的先决条件。提出了一种基于证据距离的分类器系统差异性度量,同时基于该度量提出一种多分类器系统构造方法。综合了既有差异性度量、所提新差异性度量以及在训练样本集上的分类性能等多个指标,实现了多分类器系统的有效构造。实验结果表明,所提差异性度量及多分类器系统构造方法是合理的,能有效提升融合分类精度。

关键词: 多分类器系统, 差异性度量, 证据理论, 证据距离, 多分类器融合, 分类器

Abstract: Multiple classifier systems can effectively improve the classification performance in many applications, which is why they have attracted a great deal of interest. Diversity among member classifiers is a necessary condition for improvement in classifier ensemble performance. In this paper, a novel diversity measure of multiple classifier systems is proposed based on the distance of evidence and a new approach to multiple classifier system design is presented. By using jointly the proposed diversity measure, the traditional diversity measure and the classification performance on training samples, an effective multiple classifier system can be implemented. It is experimentally shown that the proposed diversity measure and the proposed approach to multiple classifier system design are rational and effective.

Key words: multiple classifier system, diversity measure, evidence theory, distance of evidence, multiple classifier fusion, classifiers

中图分类号:

V448
TP391.4

杨艺, 韩德强, 韩崇昭. 一种基于证据距离的多分类器差异性度量[J]. 航空学报, 2012, (6): 1093-1099.

YANG Yi, HAN Deqiang, HAN Chongzhao. A Novel Diversity Measure of Multiple Classifier Systems Based on Distance of Evidence[J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2012, (6): 1093-1099.

参考文献

[1] Xu L, Krzyzak A, Suen C Y. Methods of combining multiple classifiers and their applications to handwriting recognition. IEEE Transactions on Systems, Man and Cybernetics, 1992, 22 (3): 418-435.

[2] Mashao D J, Skosan M. Combining classifier decisions for robust speaker identification. Pattern Recognition, 2006, 39(1): 147-155.

[3] Sirlantzis K, Hoque S, Fairhurst M C. Diversity in multiple classifier ensembles based on binary feature quantisation with application to face recognition. Applied Soft Computing, 2008, 8(1): 437-445.

[4] Kittler J, Hatef M, Duin R P W, et al. On combining classifiers. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1998, 20(3): 226-239.

[5] Windeatt T. Diversity measures for multiple classifier system analysis and design. Information Fusion, 2005, 6(1): 21-36.

[6] Kuncheva L I, Whitaker C J. Measures of diversity in classifier ensembles. Machine Learning, 2003, 51(2): 181-207.

[7] Shafer G. A mathematical theory of evidence. Princeton: Princeton University Press, 1967.

[8] Jousselme A L, Grenier D, Bosse E. A new distance between two bodies of evidence. Information Fusion, 2001, 2(2): 91-101.

[9] Tessem B. Approximations for efficient computation in the theory of evidence. Artificial Intelligence, 1993, 61(2): 315-329.

[10] Brown G, Wyatt J, Harris R, et al. Diversity creation methods: a survey and categorization. Information Fusion, 2005, 6(1): 5-20.

[11] Banfield R E, Hall L O, Bowyer K W, et al. Ensemble diversity measures and their application to thinning. Information Fusion, 2005, 6(1): 49-62.

[12] Partridge D, Yates W B. Engineering multiversion neural-net systems. Neural Computation, 1996, 8(4): 869-893.

[13] Bi Y X, Bell D, Guan J W. Combining evidence from classifiers in text categorization. Proceedings of 8th International Conference on KES. 2004: 521-528.

[14] Blake C L, Merz C L. UCI repository of machine learning databases. . http://www.ics.uci.edu/~mlearn/MLRepository.html.

E-mail：hkxb@buaa.edu.cn

关于我们

期刊社服务

专业学科

封面文章

友情链接

主管单位：中国科学技术协会主办单位：中国航空学会北京航空航天大学

[1]	周千里, 邓勇. 基于量子演化的信度函数概率转换[J]. 航空学报, 2022, 43(S1): 726947-726947.
[2]	尹东亮, 黄晓颖, 吴艳杰, 何有宸, 谢经伟. 基于云模型和改进D-S证据理论的目标识别决策方法[J]. 航空学报, 2021, 42(12): 324768-324768.
[3]	常琦, 杨维希, 赵恒, 孟瑶, 刘君, 高鹤明. 基于多传感器的裂纹扩展监测研究[J]. 航空学报, 2020, 41(2): 223336-223336.
[4]	王康, 史贤俊, 周绍磊, 龙玉峰, 孙美美. 基于优化SPOT和D-S证据理论的测试性验证方案[J]. 航空学报, 2019, 40(11): 223064-223064.
[5]	陈勇刚, 熊升华, 贺强, 贺元骅. 通用航空机队设备可靠性动态识别模型[J]. 航空学报, 2018, 39(3): 221706-221706.
[6]	周志文, 黄高明, 高俊. 基于压缩协作表示的辐射源识别算法[J]. 航空学报, 2016, 37(7): 2251-2258.
[7]	孙贵东, 关欣, 衣晓, 王虹. 基于云映射的多粒度语义决策属性识别[J]. 航空学报, 2015, 36(10): 3349-3358.
[8]	魏嘉利, 贾云峰, 谢树果, 吴藻菡. 航空电子系统电磁环境复杂度量化评估方法[J]. 航空学报, 2014, 35(2): 487-496.
[9]	锁斌, 曾超, 程永生, 李世玲. 认知不确定性下可靠性灵敏度分析的新指标[J]. 航空学报, 2013, 34(7): 1605-1615.
[10]	常春贺, 杨江平, 曹鹏举. 基于研制信息的测试性验证试验方案研究[J]. 航空学报, 2012, 33(11): 2057-2064.
[11]	刘准钆;程咏梅;潘泉;苗壮. 证据冲突下自适应融合目标识别算法[J]. 航空学报, 2010, 31(7): 1426-1432.
[12]	陈恬;孙健国;郝英. 基于神经网络和证据融合理论的航空发动机气路故障诊断[J]. 航空学报, 2006, 27(6): 1014-1017.
[13]	黄金;梁彦;程咏梅;潘泉;胡劲文. 基于序列图像的自动目标识别算法[J]. 航空学报, 2006, 27(1): 87-93.
[14]	龙兵;姜兴渭;宋政吉. 基于多Agent卫星遥测数据实时监测与诊断技术[J]. 航空学报, 2005, 26(6): 726-732.
[15]	邓勇;朱振福;钟山. 基于证据理论的模糊信息融合及其在目标识别中的应用[J]. 航空学报, 2005, 26(6): 754-758.

一种基于证据距离的多分类器差异性度量

A Novel Diversity Measure of Multiple Classifier Systems Based on Distance of Evidence

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价