南京大学学报(自然科学版) ›› 2023, Vol. 59 ›› Issue (6): 1034–1047.doi: 10.13232/j.cnki.jnju.2023.06.013

• • 上一篇    

基于边界域条件熵的最优尺度约简

金铭1, 陈锦坤1,2(), 孙亚超1   

  1. 1.闽南师范大学数学与统计学院,漳州,363000
    2.福建省粒计算及其应用重点实验室,闽南师范大学,漳州,363000
  • 收稿日期:2023-07-17 出版日期:2023-11-30 发布日期:2023-12-06
  • 通讯作者: 陈锦坤 E-mail:cjk99@163.com
  • 基金资助:
    国家自然科学基金(62076116);福建省自然科学基金(2020J01792)

Optimal scale reduction based on boundary domain conditional entropy

Ming Jin1, Jinkun Chen1,2(), SunYachao1   

  1. 1.School of Mathematics and Statistics,Minnan Normal University,Zhangzhou,363000,China
    2.Fujian Key Laboratory of Granular Computing and Applications,Minnan Normal University,Zhangzhou,363000,China
  • Received:2023-07-17 Online:2023-11-30 Published:2023-12-06
  • Contact: Jinkun Chen E-mail:cjk99@163.com

摘要:

由于现实世界中属性具有多层次多尺度,因此多尺度决策表的概念被提出.目前对多尺度决策表的研究大多集中在最优尺度组合上,但通过最优尺度组合得到的并不是一个真正的约简集,仍需再次进行属性约简,因此可能会导致求约简的时间较长.为此考虑利用边界域条件熵直接求最优尺度约简.首先,引入多尺度决策表中最优尺度约简的定义,给出多种最优尺度约简的定义,探讨在协调和不协调两种背景下几种最优尺度约简之间的关系.其次,给出多尺度决策表中边界域条件熵的定义,讨论边界域条件熵的若干性质以及与约简的关系.最后,给出基于边界域条件熵的最优尺度约简算法,并用实验验证该方法的有效性.

关键词: 粗糙集, 多尺度决策表, 最优尺度约简, 边界域条件熵

Abstract:

The concept of multi?scale decision tables is proposed because the attributes are with the multi?level and multi?scale in the real world. At present,most researches on multi?scale decision tables focus on optimal scale combination,but the optimal scale combination is not a real reduction set. It still needs to be reduced again. So it might lead to a higher reduction time. Therefore,the boundary domain conditional entropy is used to obtain the optimal scale reduction directly. Firstly,the definition of optimal scale reduction in multi?scale decision tables is introduced,and the definition of multiple optimal scale reductions is given. The relationship between several optimal scale reductions under the background of consistent and inconsistent is discussed. Secondly,the concept of boundary domain conditional entropy is introduced into multi?scale decision tables,and some properties of boundary domain conditional entropy and its relationship with reductions are discussed. Finally,an optimal scale reduction algorithm based on boundary domain conditional entropy is given and the effectiveness of the method is validated by experiments.

Key words: rough sets, multi?scale decision tables, optimal scale reduction, boundary domain conditional entropy

中图分类号: 

  • TP18

表1

广义多尺度决策表"

Ua11a12a21dAt(1;1)At(2;1)At(3;1)At(1;2)At(2;2)
x14T11{1,2}{1,2}{1,2}{1,2}{1,2}
x21T22{1,2}{1,2}{1,2}{1,2}{1,2}
x31T21{1,2}{1,2}{1,2}{1,2}{1,2}
x44T12{1,2}{1,2}{1,2}{1,2}{1,2}
x52F01{2}{2}{2}{2}{2}
x64T12{1,2}{1,2}{1,2}{1,2}{1,2}

表2

不同对象在不同尺度组合下的最大分布和概率分布"

UγAt(1;1)γAt(2;1)γAt(3;1)γAt(1;2)γAt(2;2)μAt(1;1)μAt(2;1)μAt(3;1)μAt(1;2)μAt(2;2)
x1D2D2D2D2D21/3,2/31/3,2/31/3,2/31/3,2/32/5,3/5
x2D1,D2D1,D2D1,D2D1,D2D21/2,1/21/2,1/21/2,1/21/2,1/22/5,3/5
x3D1,D2D1,D2D1,D2D1,D2D21/2,1/21/2,1/21/2,1/21/2,1/22/5,3/5
x4D2D2D2D2D21/3,2/31/3,2/31/3,2/31/3,2/32/5,3/5
x5D1D1D1D1D11,01,01,01,01,0
x6D2D2D2D2D21/3,2/31/3,2/31/3,2/31/3,2/32/5,3/5

表3

广义多尺度决策表"

Ua11a12a21dAt(1;1)At(2;1)At(3;1)At(1;2)At(2;2)
x14T11{1,2}{1,2}{1,2}{1,2}{1,2,3}
x24T12{1,2}{1,2}{1,2}{1,2}{1,2,3}
x32T23{2,3}{2,3}{2,3}{2,3}{1,2,3}
x43F22{2}{2}{2}{2}{2}
x52T22{2,3}{2,3}{2,3}{2,3}{1,2,3}

表4

广义多尺度决策表"

Ua11a12a13a14a21a22a31a32a33d
x16274T83995105Y
x24253F5964553N
x34753F6776374N
x48595T91993105Y
x595105T8998295Y
x6100105T92109695Y
x74853F5668295N
x85663F4154653N
x93242F8395363N

表5

预处理后的UCI实验数据集"

数据集|U|属性类别尺度
german915202(1;4;2;1;5;1;2;1;2;1;1;1;3;1;1;2;2;1;1;1)
qsarfish90864(5;4;5;5;5)
shill6321112(2;2;4;6;1;1;2;1;1;1;2)
grisonietal779112(5;4;5;9;5;5;2;1;4;1;4)
austra690142(1;4;4;1;2;3;5;1;1;5;1;1;6;6)
wifi200074(4;7;5;4;4;4;4)
raision90072(4;5;5;5;5;7;6)
hcvdat0615132(12;2;4;10;7;8;7;7;5;6;5;8;8)
frogs17952260(15;6;4;8;7;9;7;8;6;6;6;8;6;13;12;7;8;5)
auditrisk772262(1;3;7;1;7;2;1;2;3;3;1;3;7;1;7;1;1;3;5;1;5;2;7;7;4;1)
pageblocks5473105(15;3;12;11;4;4;5;17;14;12)
biascorrection75882325(4;6;4;4;6;5;7;5;2;2;2;2;11;11;12;12)

图1

不同数据集下算法的运行时间"

表6

四种算法在不同数据集下的运行时间 (s)"

数据集Lattice modelSOSCBOSCBOSR
german268.101(4)7.527 (3)3.058 (1)4.030 (2)
qsarfish127.567(4)2.325 (3)0.588 (2)0.366 (1)
shill110.627(4)25.235 (3)18.158 (2)12.552 (1)
grisonietal12375.034 (4)2.989 (3)1.285 (2)0.629 (1)
austra5189.122 (4)3.755 (3)11.509 (2)0.724 (1)
wifi3182.203 (4)5.956 (3)1.833 (2)0.983 (1)
raision1891.360 (4)0.936 (3)0.350 (2)0.180 (1)
hcvdat0*4.093 (3)2.226 (2)1.275 (1)
frogs*1191.622 (3)131.456 (2)50.495 (1)
auditrisk*5.078 (3)1.604 (2)0.702 (1)
pageblocks*11.567 (3)3.315 (2)1.756 (1)
biascorrection*1799.921 (3)205.646 (2)65.911 (1)
平均值3313.446 (4)255.084 (3)31.752 (1.92)11.634 (1.08)

表7

四种算法在部分数据集上的结果展示"

数据集Lattice modelSOSCBOSCBOSR
qsarfish(1;1;1;1;2;1)(1;1;1;1;2;1)(1;1;1;1;2;1)(1;1;1;1;2;1)
wifi(1;4;2;1;1;1;1)(1;4;2;1;1;1;1)(1;4;2;1;1;1;1)(1;4;2;1;1;1;1)
raision(1;1;1;2;1;4;1)(1;1;1;2;1;4;1)(1;1;1;2;1;4;1)(1;1;1;2;1;4;1)
hcvdat0*(1;2;2;5;7;8;1;7;5;6;5;8;7)(1;2;3;5;2;1;7;7;2;3;5;2;6)(1;3;3;5;2;1;8;8;2;3;6;2;6)
pageblocks*(1;1;12;1;1;1;1;1;14;1)(1;1;12;1;1;1;1;1;14;1)(1;1;13;1;1;1;1;1;15;1)

图2

用Bonferroni?Dunn检验比较BOSR算法与其他算法的性能"

1 Pawlak Z. Rough sets. International Journal of Computer and Information Sciences198211(5):341-356.
2 Bargiela A, Pedrycz W. Granular mappings. IEEE Transactions on Systems,Man,and Cybernetics?Part A:Systems and Humans200535(2):292-297.
3 Zadeh L A. Toward a theory of fuzzy information granulation and its centrality in human reasoning and fuzzy logic. Fuzzy Sets and Systems199790(2):111-127.
4 赵思雨,魏玲. 决策表的多种属性约简之间的关系研究. 西北大学学报(自然科学版)201747(1):18-22.
Zhao S Y, Wei L. The relationship among different attribute reduction based on a decision table. Journal of Northwest University (Natural Science Edition)201747(1):18-22.
5 周涛,陆惠玲,任海玲,等. 基于粗糙集的属性约简算法综述. 电子学报202149(7):1439-1449.
Zhou T, Lu H L, Ren H L,et al. Survey on attribute reduction algorithm of rough set. Acta Electronica Sinica202149(7):1439-1449.
6 Yuan Z, Chen H M, Xie P,et al. Attribute reduction methods in fuzzy rough set theory:An overview,comparative experiments and new directions. Applied Soft Computing2021(107):107353.
7 Leung Y, Zhang J S, Xu Z B. Clustering by scale?space filtering. IEEE Transactions on Pattern Analysis and Machine Intelligence200022(12):1396-1410.
8 Wu W Z, Leung Y. Theory and applications of granular labelled partitions in multi?scale decision tables. Information Sciences2011181(18):3878-3897.
9 Wu W Z, Leung Y. Optimal scale selection for multi?scale decision tables. International Journal of Approximate Reasoning201354(8):1107-1129.
10 魏巍,张嘉宇,陈千,等. 基于熵的多尺度决策表最优尺度选择. 海南热带海洋学院学报201825(5):61-65.
Wei W, Zhang J Y, Chen Q,et al. Entropy?based optimal scale selection in multi?scale decision tables. Journal of Hainan Tropical Ocean University201825(5):61-65.
11 郑嘉文,吴伟志,包菡,等. 基于熵的多尺度决策系统的最优尺度选择. 南京大学学报(自然科学)202157(1):130-140.
Zheng J W, Wu W Z, Bao H,et al. Entropy based optimal scale selection for multi?scale decision systems. Journal of Nanjing University (Natural Science)202157(1):130-140.
12 Hao C, Li J H, Fan M,et al. Optimal scale selection in dynamic multi?scale decision tables based on sequential three?way decisions. Information Sciences2017(415-426):213-232.
13 Chen Y S, Li J H, Li J J,et al. A further study on optimal scale selection in dynamic multi?scale decision information systems based on sequential three?way decisions. International Journal of Machine Learning and Cybernetics202213(5):1505-1515.
14 Li F, Hu B Q. A new approach of optimal scale selection to multi?scale decision tables. Information Sciences2017(381):193-208.
15 Li F, Hu B Q, Wang J. Stepwise optimal scale selection for multi?scale decision tables via attribute significance. Knowledge?Based Systems2017(129):4-16.
16 张清华,张雪秋,庞国弘. 多尺度决策系统中代价敏感的最优尺度组合. 控制与决策202136(10):2369-2378.
Zhang Q H, Zhang X Q, Pang G H. Cost?sensitive optimal scale combination in multi?scale decision systems. Control and Decision202136(10):2369-2378.
17 王金波,吴伟志. 基于证据理论的广义多尺度覆盖决策系统的最优尺度组合. 模式识别与人工智能202235(4):291-305.
Wang J B, Wu W Z. Evidence?theory?based optimal scale combinations in generalized multi?scale covering decision systems. Pattern Recognition and Artificial Intelligence202235(4):291-305.
18 Wu W Z, Leung Y. A comparison study of optimal scale combination selection in generalized multi?scale decision tables. International Journal of Machine Learning and Cybernetics202011(5):961-972.
19 牛东苒,吴伟志,李同军. 广义多尺度决策系统中基于可变精度的最优尺度组合. 模式识别与人工智能201932(11):965-974.
Niu D R, Wu W Z, Li T J. Variable precision based optimal scale combinations in generalized multi?scale decision systems. Pattern Recognition and Artificial Intelligence201932(11):965-974.
20 Cheng Y L, Zhang Q H, Wang G Y. Optimal scale combination selection for multi?scale decision tables based on three?way decision. International Journal of Machine Learning and Cybernetics202112(2):281-301.
21 Zheng J W, Wu W Z, Bao H,et al. Evidence theory based optimal scale selection for multi?scale ordered decision systems. International Journal of Machine Learning and Cybernetics202213(4):1115-1129.
22 牟琼,程云龙,吴成英,等. 不完备广义多尺度决策系统的逐步最优尺度选择. 重庆邮电大学学报(自然科学版)202335(2):210-218.
Mou Q, Cheng Y L, Wu C Y,et al. Stepwise optimal scale selection for incomplete generalized multi?scale decision systems. Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition)202335(2):210-218.
23 曾华鑫,吴伟志. 不协调广义多尺度决策系统的最优尺度组合的层次关系. 控制与决策,DOI:10.13195/j.kzyjc.2022.1648 .
Zeng H X, Wu W Z. Hierarchical relationships of optimal scale combinations in inconsistent generalized multi?scale decision systems. Control and Decision,DOI:10.13195/j.kzyjc.2022.1648 .
24 Huang Z H, Li J J, Dai W Z,et al. Generalized multi?scale decision tables with multi?scale decision attributes. International Journal of Approximate Reasoning2019(115):194-208.
25 张嘉茹,吴伟志,杨烨. 协调广义决策多尺度序信息系统的知识获取. 模式识别与人工智能202235(9):789-804.
Zhang J R, Wu W Z, Yang Y. Knowledge acquisition for consistent generalized decision multi?scale ordered information systems. Pattern Recognition and Artificial Intelligence202235(9):789-804.
26 Deng J, Zhan J M, Wu W Z. A three?way decision methodology to multi?attribute decision?making in multi?scale decision information systems. Information Sciences2021(568):175-198.
27 于子淳,吴伟志. 具有多尺度决策的信息系统的最优全局剪枝选择. 南京大学学报(自然科学版)202359(1):12-21.
Yu Z C, Wu W Z. On selection of optimal cuts in information systems with multi?scale decisions. Journal of Nanjing University (Natural Science)202359(1):12-21.
28 Cheng Y L, Zhang Q H, Wang G Y,et al. Optimal scale selection and attribute reduction in multi?scale decision tables based on three?way decision. Information Sciences2020(541):36-59.
29 She Y H, Qian Z H, He X L,et al. On generalization reducts in multi?scale decision tables. Information Sciences2020555(5):104-124.
30 金铭,陈锦坤,李进金. 基于可分离性的多尺度决策信息系统的最优尺度约简. 郑州大学学报(理学版),1-9. DOI:10.13705/j.issn.1671-6841.2022297 .
Jin M, Chen J K, Li J J. Optimal scale reduction in multi?scale decision information systems based on separability. Journal of Zhengzhou University (Natural Science Edition),1-9. DOI:10.13705/j.issn.1671-6841.2022297 .
31 程玉胜,张佑生,胡学钢. 边界条件熵的属性约简及在定性仿真中的应用. 中国科学技术大学学报200838(10):1202-1210.
Cheng Y C, Zhang Y S, Hu X G. Feature reduction based on boundary conditional entropy and its application in qualitative simulation. Journal of University of Science and Tech?nology of China200838(10):1202-1210.
32 张文修,米据生,吴伟志. 不协调目标信息系统的知识约简. 计算机学报200326(1):12-18.
Zhang W X, Mi J S, Wu W Z. Knowledge reductions in inconsistent information systems. Chinese Journal of Computers200326(1):12-18.
[1] 张义宗, 王磊, 徐阳. 属性集变化下序决策信息系统的增量属性约简算法[J]. 南京大学学报(自然科学版), 2023, 59(5): 813-822.
[2] 王冰洁, 张超, 李德玉, 马瑾男, 王渊. 基于区间二型模糊多粒度证据融合方法的钢铁行业耗能决策[J]. 南京大学学报(自然科学版), 2023, 59(4): 600-609.
[3] 徐伟华, 潘彦舟. 加权变精度直觉模糊序信息决策表的近似约简[J]. 南京大学学报(自然科学版), 2023, 59(1): 1-11.
[4] 于子淳, 吴伟志. 具有多尺度决策的信息系统的最优全局剪枝选择[J]. 南京大学学报(自然科学版), 2023, 59(1): 12-21.
[5] 章成旭, 叶绍强, 周恺卿, 欧云. 基于粗糙集和改进二进制布谷鸟搜索算法的高维数据特征选择[J]. 南京大学学报(自然科学版), 2022, 58(4): 584-593.
[6] 王文珏, 黄兵. 多尺度单值中智系统中基于优势粗糙集模型的最优尺度选择与约简[J]. 南京大学学报(自然科学版), 2022, 58(3): 495-505.
[7] 曾艺祥, 林耀进, 范凯钧, 曾伯儒. 基于层次类别邻域粗糙集的在线流特征选择算法[J]. 南京大学学报(自然科学版), 2022, 58(3): 506-518.
[8] 周悦丽, 林国平, 谢淋淋. 基于矩阵的动态局部相容粗糙集的增量方法[J]. 南京大学学报(自然科学版), 2022, 58(3): 519-531.
[9] 胡玉文, 徐久成, 张倩倩. 决策演化集的卷积预测[J]. 南京大学学报(自然科学版), 2022, 58(1): 1-8.
[10] 于子淳, 吴伟志. 用证据理论刻画协调的具有多尺度决策的信息系统的最优尺度选择[J]. 南京大学学报(自然科学版), 2022, 58(1): 71-81.
[11] 王敬前, 张小红. 基于极大相容块的不完备信息处理新方法及其应用[J]. 南京大学学报(自然科学版), 2022, 58(1): 82-93.
[12] 刘小伟, 景运革. 一种有效更新多源数据约简的增量算法[J]. 南京大学学报(自然科学版), 2021, 57(6): 1083-1091.
[13] 孙颖, 蔡天使, 张毅, 鞠恒荣, 丁卫平. 基于合理粒度的局部邻域决策粗糙计算方法[J]. 南京大学学报(自然科学版), 2021, 57(2): 262-271.
[14] 刘琼, 代建华, 陈姣龙. 区间值数据的代价敏感特征选择[J]. 南京大学学报(自然科学版), 2021, 57(1): 121-129.
[15] 郑嘉文, 吴伟志, 包菡, 谭安辉. 基于熵的多尺度决策系统的最优尺度选择[J]. 南京大学学报(自然科学版), 2021, 57(1): 130-140.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!