计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (6A): 220300202-5.doi: 10.11896/jsjkx.220300202

• 人工智能 • 上一篇    下一篇

一种新的代价敏感SVDD二类分类方法

吴崇明1, 王晓丹2, 赵振冲2   

  1. 1 西京学院商学院 西安 710123;
    2 空军工程大学防空反导学院 西安 710051
  • 出版日期:2023-06-10 发布日期:2023-06-12
  • 通讯作者: 王晓丹(afeu_w@163.com)
  • 作者简介:(afeu_w@163.com)
  • 基金资助:
    国家自然科学基金项目(61876189,61273275)

New Cost Sensitive SVDD Binary Classification Method

WU Chongming1, WANG Xiaodan2, ZHAO Zhenchong2   

  1. 1 College of Business,Xijing University,Xi’an 710123,China;
    2 College of Air and Missile Defense,Air Force Engineering University,Xi’an 710051,China
  • Online:2023-06-10 Published:2023-06-12
  • About author:WU Chongming,born in 1966,Ph.D,associate professor.His main research interests include machine learning and intelligent information processing. WANG Xiaodan,born in 1966,Ph.D,professor.Her main research interests include machine learning and pattern recognition.
  • Supported by:
    National Natural Science Foundation of China(61876189,61273275).

PDF (PC)

459

摘要： 为提升代价敏感分类性能,通过提升较高误分代价类别的学习精度来降低总误分代价,利用支持向量域描述(Support Vector Domain Description,SVDD)实现代价敏感分类,提出一种代价敏感SVDD二类分类方法CS-SVDD。该方法首先将单类SVDD拓展为二类分类SVDD,对不同类别分别构建SVDD超球体,通过误分类代价调节SVDD分类器对不同类别样本的分类精度,对误分代价高的类别进行更为精确的学习,从而降低总误分代价;对于处于两个超球体之外或覆盖区域的类别属性不明确的样本,以误分代价最小为原则定义代价敏感决策规则。在人工数据集和UCI数据集上与同类方法进行了实验比较,实验结果表明了所提方法的有效性。

关键词: 代价敏感分类, 支持向量数据描述, 支持向量

Abstract: In order to improve the performance of cost-sensitive classification,this paper improves the learning accuracy of higher misclassification cost categories to reduce the total misclassification cost,uses support vector domain description(SVDD) to reali-ze cost sensitive classification,and proposes a cost sensitive SVDD two-class classification method,CS-SVDD.This method first expands single class SVDD to two class classification SVDD,and constructs SVDD hyperspheres for different categories.by adjusting the classification accuracy of SVDD classifier for different class samples through the misclassification cost,the class with high misclassification cost can be more accurately learned,so as to reduce the total misclassification cost.For the samples with ambiguous category attributes outside the two hyperspheres or in the coverage area,cost sensitive decision rules are defined based on the principle of minimum misclassification cost.Experimental results on artificial data sets and UCI data sets show the effectiveness of the proposed method.

Key words: Cost sensitive classification, Support vector data description, Support vector

中图分类号: 

  • TP391
[1]HABBEMA J D F,HERMANS J.Cases of doubt in allocation problems[J].Biometrika,1974,61(2):313-324.
[2]ELKAN C.The Foundations of Cost-Sensitive Learning[C]//Proceedings of Proceedings of 17th International Joint Confe-rence on Artificial Intelligence.2001:973-978.
[3]DHAR S,CHERKASSKY V.Development and evaluation ofcost-sensitive universum-SVM[J].IEEE Transactions on Cybernetics,2015,45(4):806-818.
[4]YUAN G L,SUN Z W,QIN X Y,et al.Object Tracking Based on Cost Sensitive Structured SVM[J].Journal of Electronics & Information Technology,2021,43(11):3335-3341.
[5]DATTA S,DAS S.Near-Bayesian Support Vector Machines for imbalanced data classification with equal or unequal misclassification costs[J].Neural Networks,2015,70:39-52.
[6]IBÁÑEZ A,BIELZA C,LARRAÑAGA P.Cost-sensitive selective naïve Bayes classifiers for predicting the increase of the h-index for scientific journals[J].Neurocomputing,2014,135:42-52.
[7]JIANG L,LI C,WANG S.Cost-sensitive Bayesian network classifiers[J].Pattern Recognition Letters,2014,45:211-216.
[8]FREITAS A,COSTA-PEREIRA A,BRAZDIL P.Cost-sensitive decision trees applied to medical data[C]//Data Warehousing and Knowledge Discovery.Springer,2007:303-312.
[9]LI X J,ZHAO H,ZHU W.A cost sensitive decision tree algorithm with two adaptive mechanisms[J].Knowledge-Based Systems,2015,88:24-33.
[10]XIONG B Y,WANG G Y,DENG W B.Hierarchical Cost Sensi-tive Decision Tree and Its Application in Mobile Phone Replacement Prediction[J].Journal of Shandong University,2015,45(5):36-42.
[11]CHEN Y L,WUB C C,TANG K.Time-constrained cost-sensitive decision tree induction[J].Information Sciences,2016,354:140-152.
[12]SUN Y M.Cost-sensitive boosting for classification of imbal-anced data[J].Pattern Recognition,2007,40:3358-3378.
[13]XUE Y Z,WANG T.Target tracking based on cost Sensitive Adaboost[J].Chinese Journal of Image and Graphics,2016,21(5):544-555.
[14]CHENG F Y,ZHANG J,WEN C H.Cost-Sensitive Large margin Distribution Machine for classification of imbalanced data[J].Pattern Recognition Letters,2016,80:107-112.
[15]CHENG F Y,ZHANG J.Large cost-sensitive margin distribution machine for imbalanced data classification[J].Neurocomputing,2017,224:45-57.
[16]TAX D M J,DUIN R P W.Support vector domain description[J].Pattern Recognition Letters,1999,20:1191-1199.
[17]TAX D M J,DUIN R P W.Support vector domain description[J].Machine Learning,2004,54:45-66.
[18]LEE D,LEE J.Domain described support vector classifier formulti-classification problems[J].Pattern Recognition,2007,40(1):41-51.
[19]MU T T,NANDI A K.Multiclass classification based on ex-tended support vector data description[J].IEEE Trans on Systems,Man,and Cybernetics－Part B:Cybernetics,2009,39(5):1206-1216.
[20]ZHENG S F.Smoothly approximated support vector domain description[J].Pattern Recognition,2016,49:55-64.
[21]GUO Y,XIAO H,FU Q.Least square support vector data description for HRRP-based radar target recognition[J].Applied Intelligence,2017,46:365-372.
[22]WANG C D,LAI J H.Position regularized Support Vector Domain Description[J].Pattern Recognition,2013,46:875-884.
[23]ElAZAMI M.Converting SVDD scores into probability esti-mates:Application to outlier detection[J/OL].http://dx.doi.org/10.1016/j.neucom.2017.01.103.
[24]DUAN L X,XIE M Y.A new support vector data description method for machinery fault diagnosis with unbalanced datasets[J].Expert Systems With Applications,2016,64:239-246.
[25]ZHOU Y M,WU K.Fault detection of aircraft based on support vector domain description[J].Computers and Electrical Engineering,2017,61:80-94.
[26]HUANG G,CHEN H,ZHOU Z,et al.Two-class support vector data description[J].Pattern Recognition,2011,44:320-329.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发