[1]刘忠宝,张兴芹,王文莉.融合磁极效应和数据分布特征的最大间隔学习机[J].江西师范大学学报(自然科学版),2023,(06):645-651.[doi:10.16357/j.cnki.issn1000-5862.2023.06.12]
　LIU Zhongbao,ZHANG Xingqin,WANG Wenli.The Maximum Margin Learning Machine Based on Magnetic Pole Effect and Data Distribution Characteristics[J].Journal of Jiangxi Normal University:Natural Science Edition,2023,(06):645-651.[doi:10.16357/j.cnki.issn1000-5862.2023.06.12]
点击复制

融合磁极效应和数据分布特征的最大间隔学习机()
分享到：

《江西师范大学学报》（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2023年06期
页码:
645-651
栏目:
信息科学与技术
出版日期:
2023-11-25

文章信息/Info

Title:
The Maximum Margin Learning Machine Based on Magnetic Pole Effect and Data Distribution Characteristics
文章编号:
1000-5862(2023)-06-0645-07
作者:
刘忠宝1; 2; 3; 张兴芹1; 王文莉1
(1.山东外国语职业技术大学信息工程学院,山东 日照 276826; 2.北京语言大学语言智能研究院,北京 100083; 3.泉州信息工程学院软件学院,福建 泉州 362000)
Author(s):
LIU Zhongbao1; 2; 3;  ZHANG Xingqin1;  WANG Wenli1
(1.School of Information Engineering,Shandong Vocational and Technical University of International Studies,Rizhao Shandong 276826,China; 2.Institute of Language Intelligence,Beijing Language and Culture University,Beijing 100083,China; 3.School of software,Quanzhou University of Information Engineering,Quanzhou Fujian 362000,China)
关键词:
分类;  磁极效应;  数据分布;  类内离散度;  类间离散度
Keywords:
classification;  magnetic pole effect;  data distribution;  within-class scatter;  between-class scatter
分类号:
TP 391
DOI:
10.16357/j.cnki.issn1000-5862.2023.06.12
文献标志码:
A
摘要:
基于几何边界的分类方法是一种典型的智能分类方法,已有的一些方法不仅忽略数据的分布特性,而且没有考虑不同样本对分类结果的影响,因而分类精度有待于进一步提高.鉴于此,受磁极效应启发,该文提出一种新颖的融合磁极效应和数据分布特征的最大间隔学习机.该模型构造的分类超平面距离一类尽可能近,而距离另一类尽可能远,尽量地将这2类分开.该模型利用类内离散度和类间离散度来刻画数据分布特征,以期在分类决策时将数据的分布形状考虑在内.此外,模糊隶属度的引入突出了不同样本对分类结果的影响.在UCI标准数据上的比较实验表明该方法是有效的.
Abstract:
The geometric boundary based classification method is one of typical classification methods.The existed methods often neglect the data distribution and influence of different samples to the classification result,therefore,their classification accuracies can't be greatly improved.In view of this,inspired by magnetic pole effect theory,a novel classification method named maximum margin learning machine based on magnetic pole effect and data distribution characteristics(MMLM)is proposed in this paper.In this model,the hyperplane is close to one class and far away from another.The within-class scatter and between-class scatter is introduced to describe the data distribution characteristics.Meanwhile,the fuzzy membership function is used to reflect the importance of different samples.The comparative experiments on the UCI standard datasets verify the effectiveness of the proposed method.

参考文献/References:

[1] CORTES C,VAPNIK V.Support-vector networks[J].Machine Learning,1995,20(3):273-297.
[2] ZHANG Wei CHEN Junjie.Relief selection and parameter optimization for support vector machine based on mixed kernel function[J].International Journal of Performability Engineering,2018,14(2):280-289.
[3] DING Hu,XU Jinhui.Random gradient descent tree:a combinatorial approach for SVM with outliers[EB/OL].[2022-06-17].https://www.xueshufan.com/publication/2593094871.
[4] 马婷婷,杨志霞,叶俊佑.鲁棒双参数化间隔支持向量机[J].计算机工程与应用,2022,58(9):74-82.
[5] 李建民,陈慧,杨冬芹,等.改进GWO优化SVM的服务器性能预测[J].计算机工程与设计,2019,40(11):3099-3105,3163.
[6] 程凤伟,王文剑.基于近邻传输的粒度SVM算法[J].计算机科学与探索,2020,14(7):1194-1199.
[7] TAX D M J,DUIN R P W.Support vector data description[J].Machine Learning,2004,54(1):45-66.
[8] NGUYEN P,TRAN D.Repulsive-SVDD classification[C]// Proceedings of the 19th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining,May 19-22,2015,Ho Chi Minh City, Vietnam, Switzerland: Springer Cham,2015:277-288.
[9] KIM S,CHOI Y,LEE M.Deep learning with support vector data description[J].Neurocomputing,2015,165:111-117.
[10] 杨晨,王婕婷,李飞江,等.基于概率的支持向量数据描述方法[J].计算机应用,2019,39(11):3134-3139.
[11] HAO Peiyi.A new fuzzy maximal-margin spherical-structured multi-class support vector machine[EB/OL].[2022-06-17].https://ieeexplore.ieee.org/document/6890475?denied=.
[12] 陈鹏,刘爽,左莉,等.基于数据分布规律的分段组合支持向量机研究[J].微电子学与计算机,2015,32(3):94-99.
[13] 宋瑞阳,孟华,龙治国.基于数据分布特征的线性孪生支持向量机[J].计算机科学,2019,46(S1):407-411.
[14] KHANJANI-SHIRAZ R,BABAPOUR-AZAR A,HOSSEINI-NODEH Z,et al.Distributionally robust joint chance-constrained support vector machines[J].Optimization Letters,2023,17(2):299-332.
[15] BAHRAINI T,GHAZI S,YAZDI H S.Toward optimum fuzzy support vector machines using error distribution[J].Engineering Applications of Artificial Intelligence,2020,90:103545.
[16] 顾晓清,倪彤光,姜志彬,等.面向大规模噪声数据的软性核凸包支持向量机[J].电子学报,2018,46(2):347-357.
[17] 周裕群,张德生,张晓.一种改进的鲁棒模糊孪生支持向量机算法[J].计算机工程与应用,2023,59(1):140-148.
[18] 戴小路,汪廷华,周慧颖.基于加权马氏距离的模糊多核支持向量机[J].计算机科学,2022,49(S2):302-306.
[19] 刘忠宝,裴松年,杨秋翔.具有 N-S 磁极效应的最大间隔模糊分类器[J].电子科技大学学报,2016,45(2):227-232,239.

备注/Memo

备注/Memo:
收稿日期:2023-03-16
基金项目:福建省社会科学基金(FJ2022A018, FJ2021B126)资助项目.
作者简介:刘忠宝(1981—), 男, 山西太谷人,教授,博士,博士生导师,主要从事数据分析与处理的研究.E-mail:liuzb@nuc.edu.cn
更新日期/Last Update: 2023-11-25