[1]刘依菲,郭红月,刘晓东.基于样本选择的二型AFS分类方法研究[J].南京理工大学学报(自然科学版),2019,43(04):402-407.[doi:10.14177/j.cnki.32-1397n.2019.43.04.004]
　Liu Yifei,Guo Hongyue,Liu Xiaodong.Type-2 AFS classification method based on sample selection[J].Journal of Nanjing University of Science and Technology,2019,43(04):402-407.[doi:10.14177/j.cnki.32-1397n.2019.43.04.004]
点击复制

基于样本选择的二型AFS分类方法研究()
分享到：

《南京理工大学学报》（自然科学版）[ISSN:1005-9830/CN:32-1397/N]

卷:
43卷
期数:
2019年04期
页码:
402-407
栏目:
出版日期:
2019-08-24

文章信息/Info

Title:
Type-2 AFS classification method based on sample selection
文章编号:
1005-9830(2019)04-0402-06
作者:
刘依菲1; 郭红月2; 刘晓东1
1.大连海事大学 理学院,辽宁 大连116026; 2.大连海事大学 航运经济与管理学院,辽宁 大连116026
Author(s):
Liu Yifei1; Guo Hongyue2; Liu Xiaodong1
1.College of Science,Dalian Maritime University,Dalian 116026,China; 2.School of Maritime Economics and Management,Dalian Maritime University,Dalian 116026,China
关键词:
样本选择;  非平稳割点;  阈值;  分类器
Keywords:
sample selection;  unstable cut-points;  threshold;  classifier
分类号:
TP391
DOI:
10.14177/j.cnki.32-1397n.2019.43.04.004
摘要:
语义描述是聚类与分类领域的研究热点。基于公理模糊集(Axiomatic fuzzy sets,AFS)的分类算法可以模仿人类推理机理得到具有良好可解释性的类描述。在同等分类性能下,简化类描述将有助于人们较好地理解与应用分类结果。该文利用非平稳割点样本选择策略(Sample selection algorithm of unstable cut points,UCSS)和区间二型隶属函数,设计了基于非平稳割点样本选择的二型AFS(UCSS-AFS)分类方法。该分类方法在保持较好的分类准确率和语义的基础上,可以有效降低类描述的复杂度。为验证所设计方法的可实践性,该文对UCI数据库中的18个数据集进行实验。实验结果表明,UCSS-AFS分类方法在保持较好的分类准确率下所获得的规则易于理解与应用。
Abstract:
The semantic description is a hotspot in the field of clustering and classification. The classification algorithm based on axiomatic fuzzy sets(AFS)can imitate the mechanism of human reasoning to obtain the class description with sound interpretability and practicality. A concise class descriptions will help us to better understand classification results under the same classification performance. This study employs the sample selection algorithm of unstable cut points(UCSS)and interval type-2 membership function to design the type-2 AFS classification algorithm based on the sample selection of unstable cut points(UCSS-AFS). The classification method can effectively reduce the class description complexity as well as maintaining good classification accuracy and semantics. To illustrate the practicability of the designed method,this study conducts experiments on 18 data sets selected from the UCI database. Experimental results exhibit that the rules derived from the UCSS-AFS classification method are easy to understand and solve real problem with a sound classification accuracy.

参考文献/References:

[1] 蒋斌,李超英,李宗谕,等. 基于决策树分类的毫米波雷达对电力线的检测[J]. 南京理工大学学报,2017,41(1):95-99.
Jiang Bin,Li Chaoying,Li Zongyu,et al. Millimeter wave radar detection in power line based on decision tree classification[J]. Journal of Nanjing University of Science and Technology,2017,41(1):95-99.
[2]朱虹,李千目,戚湧. 一种基于改进最近邻算法的忠诚度预测方法[J]. 南京理工大学学报,2017,41(4):448-453.
Zhu Hong,Li Qianmu,Qi Yong. Loyalty prediction method base improved nearest neighbor algorithm[J]. Journal of Nanjing University of Science and Technology,2017,41(4):448-453.
[3]吴丹,顾学迈. 一种新的基于支持向量机的自动调制识别方案[J]. 南京理工大学学报,2006,30(5):569-572.
Wu Dan,Gu Xuemai. Novel scheme of automatic modulation recognition based on SVM[J]. Journal of Nanjing University of Science and Technology,2006,30(5):569-572.
[4]杨静宇,魏兴国,孙怀江. 一种快速SVM学习算法[J]. 南京理工大学学报,2003,27(5):530-535.
Yang Jingyu,Wei Xingguo,Sun Huaijiang. Fast SVM learning algorithm[J]. Journal of Nanjing University of Science and Technology,2003,27(5):530-535.
[5]黄宇扬,董明刚,敬超. 面向最近邻分类的遗传实例选择算法[J]. 计算机应用,2018,38(11):3112-3118.
Huang Yuyang,Dong Minggang,Jing Chao. Genetic instance selection algorithm for K-nearest neighbor classifier[J]. Journal of Computer Applications,2018,38(11):3112-3118.
[6]Malhat M,Menshawy M E,Mousa H,et al. Improving instance selection methods for big data classification[C]//Computer Engineering Conference(ICENCO)2018 14th International. Cairo,Egypt:IEEE,2018:90-95.
[7]Hart P. The condensed nearest neighbor rule[J]. IEEE Transactions on Information Theory,2003,14(3):515-516.
[8]Ritter G L,Woodruff H B,Lowry S R,et al. An algorithm for a selective nearest neighbor decision rule[J]. IEEE Transactions on Information Theory,1975,21(6):665-669.
[9]Gates G W. The reduced nearest neighbor rule[J]. IEEE Transactions on Information Theory,1972,18(3):431-433.
[10]Wilson D L. Asymptotic properties of nearest neighbor rules using edited data[J]. IEEE Transactions on Systems,Man and Cybernetics,1972,SMC-2(3):408-421.
[11]Aha D W,Kibler D,Albert M K. Instance-based learning algorithms[J]. Machine Learning,1991,6(1):37-66.
[12]王熙照,邢胜,赵士欣. 基于非平稳割点的大数据分类样例选择[J]. 模式识别与人工智能,2016,29(9):780-789.
Wang Xizhao,Xing Sheng,Zhao Shixin. Unstable cut-points based sample selection for large data classification[J]. Pattern Recognition and Artificial Intelligence,2016,29(9):780-789.
[13]Liu W,Liu X D. The framework of axiomatics fuzzy sets based fuzzy classifiers[J]. Journal of Industrial and Management Optimization,2017,4(3):581-609.
[14]张盼盼. 基于AFS的区间二型隶属函数构建方法及应用[D]. 大连:大连海事大学理学院,2018.
[15]Liu X D,Pedrycz W. AFS fuzzy classifiers[J]. Studies in Fuzziness and Soft Computing,2009,244:423-489.
[16]Liu X D. A new mathematical axiomatic system of fuzzy sets and systems[J]. Journal of Fuzzy Mathematics,1995,3:559-560.
[17]Liu X D. The development of AFS theory under probability theory[J]. International Journal of Information and Systems Sciences,2007,3(2):326-348.

备注/Memo

备注/Memo:
收稿日期:2019-04-28 修回日期:2019-05-21
基金项目:国家自然科学基金(61803065); 中央高校基本科研业务费专项基金(3132019175); 辽宁省社会科学规划基金项目(L18DGL010)
作者简介:刘依菲(1994-)女,硕士,主要研究方向:数据挖掘、模糊数学,E-mail:lyf1994@dlmu.edu.cn; 通讯作者:郭红月(1989-)女,博士,主要研究方向:粒计算、预测与决策,E-mail:hyguo@dlmu.edu.cn。
引文格式:刘依菲,郭红月,刘晓东. 基于样本选择的二型AFS分类方法研究[J]. 南京理工大学学报,2019,43(4):402-407.
投稿网址:http://zrxuebao.njust.edu.cn
更新日期/Last Update: 2019-09-30