[1]寇振宇,杨绪兵,张福全,等.L1范数最大间隔分类器设计[J].南京师范大学学报(自然科学版),2018,41(04):59.[doi:10.3969/j.issn.1001-4616.2018.04.010]
　Kou Zhenyu,Yang Xubing,Zhang Fuquan,et al.Design of L1 Norm Maximum Margin Classifier[J].Journal of Nanjing Normal University(Natural Science Edition),2018,41(04):59.[doi:10.3969/j.issn.1001-4616.2018.04.010]
点击复制

L1范数最大间隔分类器设计()
分享到：

《南京师范大学学报》（自然科学版）[ISSN:1001-4616/CN:32-1239/N]

卷:
第41卷
期数:
2018年04期
页码:
59
栏目:
·数学与计算机科学·
出版日期:
2018-12-31

文章信息/Info

Title:
Design of L1 Norm Maximum Margin Classifier
文章编号:
1001-4616(2018)04-0059-06
作者:
寇振宇1; 杨绪兵1; 张福全1; 杨红鑫1; 许等平2
(1.南京林业大学信息科学技术学院,南京 210037)(2.国家林业局调查规划设计院,北京 100714)
Author(s):
Kou Zhenyu1; Yang Xubing1; Zhang Fuquan1; Yang Hongxin1; Xu Dengping2
(1.College of Information Science and Technology,Nanjing Forestry University,Nanjing 210037,China)(2.Survey & Planning Institute of State Forestry Administration,Beijing 100714,China)
关键词:
L1范数; 支持向量机; 间隔; 线性规划
Keywords:
L1 norm; support vector machine; margin; linear programming
分类号:
TP391
DOI:
10.3969/j.issn.1001-4616.2018.04.010
文献标志码:
A
摘要:
以L1范数为例,设计了一个L1范数的大间隔分类器L1MMC(L1-norm Maximum Margin Classifier),主要特点如下:(1)间隔由L1范数的点到平面距离解析表示;(2)该分类器与SVM一样,通过最大化L1间隔,达到同时最小化经验风险和结构风险的目的;(3)只需要通过线性规划进行求解,避免了SVM的二次规划问题;(4)分类精度达到甚至超过SVM. 最后,在人工数据和国际标准UCI数据集上,验证了该方法的有效性.
Abstract:
L1 norm is taken as an example to design an L1 norm L1MMC(L1-norm Maximum Margin Classifier). The main features are as follows:(1)The interval is represented by the point-to-plane distance analysis of the L1 norm;(2)This classifier,like SVM,maximizes the L1 interval to minimize the risk of both empirical and structural risks;(3)Only need to be solved through linear programming to avoid the quadratic programming problem of SVM;(4)Classification accuracy reaches or even exceeds SVM. Finally,on the artificial data and the international standard UCI data set,verify the effectiveness of the method.

参考文献/References:

[1] HUANG K,YANG H,KING I,et al. Maxi-min margin machine:learning large margin classifiers locally and globally[J]. IEEE transactions on neural networks,2008,19(2):260-272.
[2]VAPNIK V. The nature of statistical learning theory[C]//Conference on artificial intelligence. Montreal,Canada:Springer-Verlag,1995:988-999.
[3]SUYKENS J A K,VANDEWALLE J. Least squares support vector machine classifiers[J]. Neural processing letters,1999,9(3):293-300.
[4]FUNG G. Proximal support vector machine classifier[J]. Machine learning,2005,59(1/2):77-97.
[5]MANGASARIAN O L,WILD E W. Multisurface proximal support vector machine classification via generalized eigenvalues[J]. IEEE transactions on pattern analysis & machine intelligence,2006,28(1):69-74.
[6]HUANG K,ZHENG D,KING I,et al. Arbitrary norm support vector machines[J]. Neural computation,2014,21(2):560-582.
[7]GU B,SHENG V S. A robust regularization path algorithm for γ-support vector classification[J]. IEEE transactions on neural networks and learning systems,2017,28(5):1241-1248.
[8]RASTOGI R,SAIGAL P,CHANDRA S. Angle-based twin parametric-margin support vector machine for pattern classification[J]. Knowledge-based systems,2018,139:64-77.
[9]NAN S,SUN L,CHEN B,et al. Density-dependent quantized least squares support vector machine for large data sets[J]. IEEE transactions on neural networks and learning systems,2017,28(1):94-106.
[10]TIPPING M E. Sparse Bayesian learning and the relevance vector machine[J]. Journal of machine learning research,2001,1(3):211-244.
[11]ZHENG W,LIN Z,WANG H. L1-norm kernel discriminant analysis via Bayes error bound optimization for robust feature extraction[J]. IEEE transactions on neural networks and learning systems,2014,25(4):793-805.
[12]WANG L,SHEN X. On L1-norm multiclass support vector machines:methodology and theory[J]. Journal of the American statistical association,2007,102(478):583-594.
[13]YAN H,YE Q,YU D J,et al. Least squares twin bounded support vector machines based on L1-norm distance metric for classification[J]. Pattern recognition,2018,74:434-447.
[14]ZHU J,ROSSET S,TIBSHIRANI R,et al. 1-norm support vector machines[C]//Advances in neural information processing systems. Vancouver,Canada:MIT Press,2003:49-56.
[15]WANG L,SHEN X,ZHENG Y F. On L1-norm multi-class support vector machines[C]//International conference on machine learning and applications. Orlando,USA:IEEE Computer Society,2006:83-88.
[16]KUJALA J,AHO T,ELOMAA T. A walk from 2-norm SVM to 1-norm SVM[C]//Ninth IEEE international conference on data mining. Miami,USA:IEEE Computer Society,2009:836-841.
[17]杨绪兵,顾一凡,陈松灿,等. L1投影的解析计算方法[J]. 南京大学学报(自然科学版),2017,53(3):476-482.
[18]杨绪兵,王一雄,陈斌. 马氏度量学习中的几个关键问题研究及几何解释[J]. 南京大学学报(自然科学版),2013,49(2):133-141.
[19]YEN E H,ZHONG K,HSIEH C J,et al. Sparse linear programming via primal and dual augmented coordinate descent[C]//International conference on neural information processing systems. Montreal,Canada:MIT Press,2015:2368-2376.

相似文献/References:

[1]朱志宾,丁世飞.基于TWSVM的图像分类[J].南京师范大学学报(自然科学版),2014,37(03):8.
　Zhu Zhibin,Ding Shifei.Image Classification Based on Twin Support Vector Machines[J].Journal of Nanjing Normal University(Natural Science Edition),2014,37(04):8.
[2]王 征,李皓月,许洪山,等.基于卷积神经网络和SVM的中国画情感分类[J].南京师范大学学报(自然科学版),2017,40(03):74.[doi:10.3969/j.issn.1001-4616.2017.03.011]
　Wang Zheng,Li Haoyue,Xu Hongshan,et al.Chinese Painting Emotion Classification Based onConvolution Neural Network and SVM[J].Journal of Nanjing Normal University(Natural Science Edition),2017,40(04):74.[doi:10.3969/j.issn.1001-4616.2017.03.011]
[3]汤嘉立,朱广萍,杜卓明.支持向量机多特征分类学习的超分辨率复原[J].南京师范大学学报(自然科学版),2018,41(03):28.[doi:10.3969/j.issn.1001-4616.2018.03.005]
　Tang Jiali,Zhu Guangping,Du Zhuoming.Super-resolution Restoration Algorithm Based onSVM Multi-figure Classification Learning[J].Journal of Nanjing Normal University(Natural Science Edition),2018,41(04):28.[doi:10.3969/j.issn.1001-4616.2018.03.005]

备注/Memo

备注/Memo:
收稿日期:2018-08-16.
基金项目:国家自然科学基金(31670554、61871444)、江苏省自然科学基金(BK20161527、BK20171453).
通讯联系人:杨绪兵,博士,副教授,研究方向:模式识别,机器学习. Email:xbyang@njfu.edu.cn
更新日期/Last Update: 2018-12-30