融合磁极效应和数据分布特征的最大间隔学习机  

The Maximum Margin Learning Machine Based on Magnetic Pole Effect and Data Distribution Characteristics

作　　者：刘忠宝 张兴芹 王文莉 LIU Zhongbao;ZHANG Xingqin;WANG Wenli(School of Information Engineering,Shandong Vocational and Technical University of International Studies,Rizhao Shandong 276826,China;Institute of Language Intelligence,Beijing Language and Culture University,Beijing 100083,China;School of software,Quanzhou University of Information Engineering,Quanzhou Fujian 362000,China)

机构地区：[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

出　　处：《江西师范大学学报（自然科学版）》2023年第6期645-651,共7页Journal of Jiangxi Normal University(Natural Science Edition)

基　　金：福建省社会科学基金(FJ2022A018,FJ2021B126)资助项目。

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.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.

关 键 词：分类 磁极效应 数据分布 类内离散度 类间离散度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]