南京大学学报(自然科学版) ›› 2023, Vol. 59 ›› Issue (1): 1–11.doi: 10.13232/j.cnki.jnju.2023.01.001

• •    下一篇

加权变精度直觉模糊序信息决策表的近似约简

徐伟华(), 潘彦舟   

  1. 西南大学人工智能学院,重庆,400715
  • 收稿日期:2022-09-26 出版日期:2023-01-31 发布日期:2023-03-01
  • 通讯作者: 徐伟华 E-mail:chxuwh@gmail.com
  • 基金资助:
    国家自然科学基金(61976245)

Approximate reduction in weighted variable precision intuitionistic fuzzy ordered decision table

Weihua Xu(), Yanzhou Pan   

  1. College of Artificial Intelligence,Southwest University,Chongqing,400715,China
  • Received:2022-09-26 Online:2023-01-31 Published:2023-03-01
  • Contact: Weihua Xu E-mail:chxuwh@gmail.com

摘要:

以直觉模糊信息表为背景,利用粗糙集和模糊集,旨在筛除信息表中冗余的属性,提出获取决策规则的近似约简方法.首先,通过在直觉模糊集中引入带权重评分函数来定义加权直觉模糊序关系;进一步,为了提高模型分类的容错率,结合变精度粗糙集模型构建加权变精度直觉模糊序决策信息表;接着,在该决策表中提出上、下近似约简的判定定理和可辨识矩阵,进而生成两种求解上、下近似约简的方法;最后,通过具体案例和数值实验分析验证了该方法的有效性.

关键词: 变精度粗糙集, 带权重评分函数, 近似约简, 可辨识矩阵, 直觉模糊集

Abstract:

Aiming at removing irrelevant and redundant attributes,this paper proposes a method of approximate reduction to better require decision rules in intuitionistic fuzzy decision tables. The weighted system is defined via the introduction of a weighted score function in intuitionistic fuzzy sets. Additionally,variable precision rough sets are utilized for a better tolerance to misclassify. Hence,the weighted variable precision intuitionistic fuzzy sets are defined. On the basis of the constructed system,we give conceptions of the judgment theorem and identification matrix of both lower and upper approximate reduction,by which two approaches of reduction are put forward. Finally,a concrete example and numerical tests are used to illustrate the effectiveness of the proposed method.

Key words: variable precision rough sets, weighted score function, approximate reduction, identification matrix, intuitionistic fuzzy sets

中图分类号: 

  • TP18

表1

加权变精度直觉模糊序决策信息表"

Ua1a2a3d
x10.4,0.20.5,0.30.4,0.1A
x20.9,0.10.6,0.20.6,0.4B
x30.3,0.50.3,0.40.5,0.2C
x40.6,0.30.2,0.60.8,0.1B
x50.8,0.10.7,0.20.5,0.3A
x60.9,00.4,0.40.7,0C

表2

基于加权变精度直觉模糊序决策信息表(表1)的得分函数"

Ua1a2a3d
x10.140.190.16A
x20.510.280.24B
x30.010.030.21C
x40.26-0.080.44B
x50.440.350.19A
x60.530.10.39C

表3

表1的下近似辨识矩阵"

Ux1x2x3x4x5x6
x1a1,2a2
x2a1,2,3a2
x3
x4a1,3a3
x5a1,2,3a2a1,2a1,2a2
x6

表4

表1的上近似辨识矩阵"

Ux1x2x3x4x5x6
x1
x2a1,2,3a1,3
x3a3a3
x4a1,3a3
x5
x6a1,3a1,3a1,2,3a1,2a1,3

表5

实验中使用的数据集概述"

数据集简称样本数属性数分类数
WineWine178133
SeedsSeeds21073

Heart Failure

Clinical records

Heart299132
Forest FiresForest517132

Wisconsin Diagnostic

Breast Cancer

Wdbc569312
Australian CreditAust690142
German CreditGerman1000202
Maternal HealthHealth101473
CardiotocographyCard2093213

表6

不同算法在KNN和SVM上的分类精度"

数据集KNNSVM
VDRVURMRVDRVURMR
Average85.70%±6.96%85.50%±5.70%83.19%±6.81%86.01%±6.75%84.61%±6.62%84.47%±5.89%
Wine98.33%±5.00%96.33%±7.37%96.00%±0.6496.33%±7.37%96.00%±8.00%96.00%±0.64%
Seeds94.05%±7.32%95.48±6.94%92.38%±7.66%95.48%±6.94%96.90%±6.21%91.91%±8.12%
Heart82.22%±14.2373.33%±7.37%81.11%±12.22%76.67%±13.56%71.11%±7.37%83.33%±12.42%
Forest91.96%±6.59%97.14%±5.71%97.14%±5.71%94.46%±6.79%92.86%±9.58%98.75%±3.75%
Wdbc94.74%±5.55%94.12%±5.88%93.01%±6.29%95.88%±4.59%92.94%±5.76%93.56%±5.55%
Aust87.45%±5.37%87.48%±5.37%70.67%±10.42%85.95%±9.47%82.55%±10.54%73.02%±8.64%
German70.33%±5.47%73.33%±2.98%72.00%±8.59%73.00%±1.00%73.00%±1.00%72.00%±3.27%
Health64.62%±8.83%64.26%±6.03%61.63%±6.57%67.86%±7.46%67.20%±7.59%67.20%±7.59%
Card87.61%±4.20%88.09%±3.67%84.81%±3.17%88.41%±3.57%88.87%±3.52%47.49%±2.97%

表7

不同算法在BayesNet和RandomTree上的分类精度"

数据集BayesNetRandomTree
VDRVURMRVDRVURMR
Average83.72%±6.14%83.68%±6.69%81.76%±7.79%89.83%±7.18%89.82%±8.26%84.85%±5.89%
Wine95.00%±4.99%93.00%±7.37%90.98%±2.18%98.00%±8.19%98.33%±7.37%92.67%±1.74%
Seeds95.48%±6.94%96.9%±7.9%97.14%±5.71%96.67%±7.64%95.00%±8.12%93.81%±7.62%
Heart83.33%±12.62%82.22%±10.48%87.78%±20.76%89.96%±3.18%88.89%±7.45%82.22%±11.60%
Forest97.14%±5.71%97.14%±5.71%94.46%±6.55%98.75%±9.01%98.57%±9.46%95.71%±6.8%
Wdbc95.88%±3.90%95.39%±4.88%90.03%±5.44%96.47%±5.29%97.06%±6.34%91.80%±4.88%
Aust82.67%±7.03%82.67%±7.05%72.55%±10.18%88.40%±6.69%87.43%±5.98%75.41%±5.73%
German67.47%±4.16%69.33%±5.78%68.00%±8.14%73.33%±12.83%76.33%±14.59%70.33%±5.59%
Health61.69%±6.28%62.01%±7.51%62.01%±6.96%74.39%±6.67%74.73%±9.30%76.42%±9.30%
Card74.59%±3.63%74.44%±3.48%72.89%±4.18%92.47%±5.12%91.20%±5.74%84.59%±4.73%
1 Atanassov K T. Intuitionistic fuzzy sets. Fuzzy Sets and Systems198620(1):87-96.
2 Zadeh L A. Fuzzy sets. Information and Control19658(3):338-353.
3 那日萨,孔茸,高欢. 基于深度学习的直觉模糊集隶属度确定方法. 运筹与管理202231(2):92-98.
Na R S, Kong R, Gao H. Deep learning?based determination method for membership degree in intuitionistic fuzzy sets. Operations Research and Management Science202231(2):92-98.
4 Deng Y, Ren Z Q, Kong Y Y,et al. A hierarchical fused fuzzy deep neural network for data classification. IEEE Transactions on Fuzzy Systems201725(4):1006-1012.
5 Zhang D, Li Y L, Wu C. An extended TODIM method to rank products with online reviews under intuitionistic fuzzy environment. Journal of the Operational Research Society202071(2):322-334.
6 戴文战,王琪. 基于PCNN与IFS的可见光与红外图像融合方法. 光电子·激光202031(7):738-744.
Dai W Z, Wang Q. Research on fusion method of visible and infrared image based on PCNN and IES. Journal of Optoelectronics · Laser202031(7):738-744.
7 Pawlak Z. Rough sets. International Journal of Computer & Information Sciences198211(5):341-356.
8 Greco S, Matarazzo B, Slowinski R. Rough approximation of a preference relation by dominance relations. European Journal of Operational Research1999117(1):63-83.
9 Yao Y Y. Relational interpretations of neighborhood operators and rough set approximation operators. Information Sciences1998111(1-4):239-259.
10 Dubois D, Prade H. Rough fuzzy sets and fuzzy rough sets. International Journal of General Systems199017(2-3):191-209.
11 Chen X W, Xu W H. Double?quantitative multigranulation rough fuzzy set based on logical operations in multi?source decision systems. International Journal of Machine Learning and Cybernetics202213(4):1021-1048.
12 Che X Y, Mi J S, Chen D G. Information fusion and numerical characterization of a multi?source infor?mation system. Knowledge?Based Systems2018,145:121-133. DOI:10.1016/j.knosys. 2018.01.008 .
13 王江荣,黄建华,罗资琴,等. 基于粗糙集的Logistic回归模型在矿井突水模式识别中的应用. 煤田地质与勘探201543(6):70-74.
Wang J R, Huang J H, Luo Z Q,et al. Application of logistic regression model based on rough set in recognition of mine water inrush pattern. Coal Geology & Exploration201543(6):70-74.
14 Sang B B, Chen H M, Yang L,et al. Incremental attribute reduction approaches for ordered data with time?evolving objects. Knowledge?Based Systems2021(212):106583. DOI:10.1016/j.knosys. 2020. 106583 .
15 林冰雁,徐伟华,杨倩. 带偏好度量的直觉模糊序决策信息系统的部分一致约简. 计算机科学201845(1):148-151,187.
Lin B Y, Xu W H, Yang Q. Partially consistent reduction in intuitionistic fuzzy ordered decision information systems with preference measure. Computer Science201845(1):148-151,187.
16 Chen Y M, Zeng Z Q, Zhu Q X,et al. Three?way decision reduction in neighborhood systems. Applied Soft Computing2016(38):942-954. DOI:10.1016/j.asoc.2015.10.059 .
17 龙柄翰,徐伟华,张晓燕. 不协调目标信息系统中基于改进差别信息树的分布属性约简. 计算机科学201946(S1):115-119.
Long B H, Xu W H, Zhang X Y. Distribution attribute reduction based on improved discernibility information tree in inconsistent system. Computer Science201946(S1):115-119.
18 Ziarko W. Variable precision rough set model. Journal of Computer and System Sciences199346(1):39-59.
19 Chen Y Y, Chen Y M. Feature subset selection based on variable precision neighborhood rough sets. International Journal of Computational Intelligence Systems202114(1):572-581.
20 Chen P P, Lin M L, Liu J H. Multi?label attribute reduction based on variable precision fuzzy neighborhood rough set. IEEE Access2020(8):133565-133576. DOI:10.1109/ACCESS.2020. 3010314 .
21 陈子春,秦克云. 区间值信息系统在变精度相容关系下的属性约简. 计算机科学200936(3):163-166.
Chen Z C, Qin K Y. Attribute reduction of interval?valued information system based on variable precision tolerance relation. Computer Science200936(3):163-166.
22 Hu M, Tsang E C C, Guo Y T,et al. A novel approach to attribute reduction based on weighted neighborhood rough sets. Knowledge?Based Systems2021(220):106908.
23 Xie X J, Qin X L, Yu C Q,et al. Test?cost?sensitive rough set based approach for minimum weight vertex cover problem. Applied Soft Computing2018(64):423-435. DOI:10.1016/j.asoc.2017. 12.023 .
24 Aggarwal M. Probabilistic variable precision fuzzy rough sets. IEEE Transactions on Fuzzy Systems201624(1):29-39.
25 徐伟华,张文修. 基于优势关系下的协调近似空间. 计算机科学200532(9):164-165.
Xu W H, Zhang W X. Consistent approximation spaces based on dominance relations. Computer Science200532(9):164-165.
26 徐泽水. 直觉模糊信息集成理论及应用. 北京:科学出版社,2008.
27 徐伟华. 序信息系统与粗糙集. 北京:科学出版社,2013.
28 张文修,梁怡,吴伟志. 信息系统与知识发现. 北京:科学出版社,2003.
29 陈德刚,徐伟华,李金海,等. 粒计算基础教程. 北京:科学出版社,2019.
[1] 徐伟华, 孔子默, 陈曜琦. 含属性加权模糊序决策信息系统的近似约简[J]. 南京大学学报(自然科学版), 2022, 58(2): 255-263.
[2] 敬思惠,秦克云*. 决策系统基于特定决策类的上近似约简[J]. 南京大学学报(自然科学版), 2018, 54(4): 804-.
[3] 张振华1*,林小龙1,甘穗福2,袁申国3,胡 勇4. 一类带参数直觉模糊知识度量方法[J]. 南京大学学报(自然科学版), 2017, 53(6): 1100-.
[4]  石素玮1*,谭安辉2.  基于诱导覆盖的粗糙直觉模糊集模型[J]. 南京大学学报(自然科学版), 2017, 53(5): 947-.
[5] 张振华1*,胡 勇2,严玉清3. 一类基于坐标变换的带参数直觉模糊距离构造方法[J]. 南京大学学报(自然科学版), 2017, 53(3): 462-.
[6] 郭郁婷,李进金,李克典,郭玉龙. 多粒度覆盖粗糙直觉模糊集模型[J]. 南京大学学报(自然科学版), 2015, 51(2): 438-446.
[7] 程玉胜1**,江效尧2,胡林生2
.  基于粗糙集理论的协调集及其决策树构造*[J]. 南京大学学报(自然科学版), 2012, 48(6): 790-796.
[8]  赵荣泳 1 , 李翠玲 2 ** , 高晓康 3 , 王昭云 4
.  电子镇流器故障诊断的变精度粗糙集模型*
[J]. 南京大学学报(自然科学版), 2010, 46(5): 494-500.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!