| 3 | 0 | 3 |
| 下载次数 | 被引频次 | 阅读次数 |
海量数据的约简可以有效节省算力.为解决动态环境下缺失型数据的属性约简问题，本文依据可比粒度距离提出一种新型双邻域粗糙集模型.首先给出依赖差异度的不完备距离刻画机制以及新的可比粒度距离函数，并构建双邻域粗糙集模型；然后由此建立属性依赖度及增量式约简算法，以应对现实数据集中实时更新变化的复杂数据；最后通过多个UCI数据集进行仿真实验.仿真结果表明，本文所提出的增量式约简算法在处理缺失型及动态数据时能保证显著的约简效率和分类精度.
Abstract：The approximation of massive data can effectively save the arithmetic power, in order to solve the attribute approximation problem of missing-type data in dynamic environment, a new Tri-partition neighborhood rough sets model is proposed based on comparable granularity distance. Firstly, the incomplete distance carving mechanism depending on the degree of difference and the new comparable granularity distance function are given, and the Tri-partition neighborhood rough sets model is constructed; then the attribute dependency and simplification algorithms are established from this to cope with the complex data that are updated and changed in real time in real datasets; finally, simulation experiments are carried out through multiple UCI datasets. The results show that the proposed Incremental Feature Selection can guarantee significant approximation efficiency and classification accuracy when dealing with missing and dynamic data.
[1]Cheng X Q,Mei H,Zhao W,et al.Data science and computing intelligence:concept,paradigm,and opportunities[J].Bulletin of Chinese Academy of Sciences,2020,35(12):1470-1481.
[2]胡清华，于达仁，谢宗霞.基于邻域粒化和粗糙逼近的数值属性约简[J].软件学报，2008,3:640-649.
[3]周艳红，张贤勇，莫智文.粒化单调的条件邻域熵及其相关属性约简[J].计算机研究与发展，2018,55(11):2395-2405.
[4]Yao Y Y,Qi J J,Wei L.Three-way granular computing,rough sets,and formal concept analysis[J].International Journal of Approximate Reasoning,2020,116(11):106-125.
[5]Song Y,Lin H,Li Z W.Outlier detection in a multiset-valued information system based on rough set theory and granular computing[J].Information Sciences,2024,657:119950.
[6]Su X,Yuan Z,Chen B,et al.Detecting anomalies with granular-ball fuzzy rough sets[J].Information Sciences,2024,678:121016.
[7]Chen Y,Cheng Y,Luo B,et al.A novel attribute reduction algorithm based on granular sequential three-way decision[J].Information Sciences,2025,694:121691-121691.
[8]Salehi F,Keyvanpour M R,Sharifi A.SMKFC-ER:semi-supervised multiple kernel fuzzy clustering based on entropy and relative entropy[J].Information Sciences,2021,547(8):667-688.
[9]何松华，康婵娟，鲁敏，等.基于邻域组合测度的属性约简方法[J].控制与决策，2016,31(7):1225-1230.
[10]Zhao H,Qin K.Mixed feature selection in incomplete decision table[J].Knowledge-Based Systems,2014,57(12):181-190.
[11]Wang C,Fan H,Wu T.Novel rough set theory-based method for epistemic uncertainty modeling,analysis and applications[J].Applied Mathematical Modelling,2023,113(9):456-474.
[12]Tong S,Sun B,Zhang L,et al.An approach of multi-criteria group decision making with incomplete information based on formal concept analysis and rough set[J].Expert Systems With Applications,2024,248:123364.
[13]王标，卫红权，王凯，等.基于粗糙集的不完备谣言信息系统的知识获取与决策[J].信息安全学报，2024,9(2):19-35.
[14]盛魁，王伟，卞显福，等.混合数据的邻域区分度增量式属性约简算法[J].电子学报，2020,48(4):682-696.
[15]Yang C,Ge H,Li L,et al.A unified incremental reduction with the variations of the object for decision tables[J].Soft Computing,2019,23(15):6407-6427.
[16]Zhao J,Ling Y,Huang F,et al.Incremental feature selection for dynamic incomplete data using sub-tolerance relations[J].Pattern Recognition,2024,148:110125.
[17]Yue X,Chen Y,Miao D,et al.Tri-partition neighborhood covering reduction for robust classification[J].International Journal of Approximate Reasoning,2017,83(11):371-384.
[18]Chen Y,Qin N,Li W,et al.Granule structures,distances and measures in neighborhood systems[J].Knowledge-Based Systems,2018,165(11):268-281.
[19]姚晟，汪杰，徐风，等.不完备邻域粗糙集的不确定性度量和属性约简[J].计算机应用，2018,38(1):97-103.
[20]郑颖春，郭玲.一种改进的基于知识粒度的增量属性约简算法[J].小型微型计算机系统，2025,46(1):128-134.
基本信息:
DOI：10.19943/j.2095-3070.jmmia.2025.04.02
中图分类号:TP18
引用信息:
[1]丁冬,骆公志.基于可比粒度距离的双邻域粗糙集增量式约简算法[J].数学建模及其应用,2025,14(04):10-20.DOI:10.19943/j.2095-3070.jmmia.2025.04.02.
基金信息:
国家自然科学基金(72171124); 江苏高校哲学社会科学研究重大项目(2021SJZDA129); 江苏省研究生科研创新计划项目(KYCX24＿1091)