基于粒球邻域粗糙集的三支属性约简

Three-way attribute reduction based on granular-ball neighborhood rough set

摘要

本文提出了一种基于粒球邻域粗糙集(granular-ball neighborhood rough Set, GBNRS)的三支属性约简算法.首先,构建一种基于粒球样本分布的纯度度量准则,通过同时考虑粒球内部样本分布和粒球间的分布,对传统度量进行改进,以更准确地评估属性集合对决策分类的贡献.在此基础上,引入三支决策机制,将属性划分为核心、候选与冗余3类,实现分层筛选与渐进式约简,从而降低误判风险并提升约简质量.最后,通过在12个UCI公共数据集上与多种方法进行对比实验,结果表明所提算法能够在减少约简长度的同时保持较好的分类性能.

Abstract

This paper proposes a three-way attribute reduction algorithm based on GBNRS. First, a purity measure based on granular-ball sample distribution is constructed. By simultaneously considering the sample distribution within granular balls and the distribution among granular balls, the traditional measure is improved to evaluate the contribution of attribute subsets to decision classification more accurately. On this basis, a three-way decision mechanism is introduced to divide attributes into core, candidate, and redundant attributes, enabling hierarchical screening and progressive reduction, thereby reducing the risk of misjudgment and improving reduction quality. Finally, comparative experiments with multiple methods are conducted on 12 UCI public datasets. The results show that the proposed algorithm can reduce the reduct length while maintaining good classification performance.

关键词

粒球计算 / 粒球邻域粗糙集 / 属性约简 / 三支决策

Key words

granular-ball computing / granular-ball neighborhood rough set / attribute reduction / three-way decision

参考文献

[1] Pawlak Z . Rough sets[J]. International Journal of Computer & Information Sciences, 1982, 11(5): 341-356.

[2] Sun L, Xu F, Ding W P, et al. AFIFC: Adaptive fuzzy neighborhood mutual information-based feature selection via label correlation[J]. Pattern Recognition, 2025, 164: 111577.

[3] Ali G, Afzal M, Asif M, et al. Attribute reduction approaches under interval-valued q-rung orthopair fuzzy soft framework[J]. Applied Intelligence, 2022, 52(8): 8975-9000.

[4] Sewwandi M A N D, Li Y F, Zhang J L . A class-specific feature selection and classification approach using neighborhood rough set and K-nearest neighbor theories[J]. Applied Soft Computing, 2023, 143: 110366.

[5] Hu Q H, Yu D R, Liu J F, et al. Neighborhood rough set based heterogeneous feature subset selection[J]. Information Sciences, 2008, 178(18): 3577-3594.

[6] Hu M, Tsang E C C, Guo Y T, et al. A novel approach to attribute reduction based on weighted neighborhood rough sets[J]. Knowledge-Based Systems, 2021, 220: 106908.

[7] Xie Q, Zhang Q H, Xia S Y, et al. GBG: a fast and stable granular ball generation method for classification[J]. IEEE Transactions on Emerging Topics in Computational Intelligence, 2024, 8(2): 2022-2036.

[8] Cheng D D, Liu S S, Xia S Y, et al. Granular-ball computing-based manifold clustering algorithms for ultra-scalable data[J]. Expert Systems with Applications, 2024, 247: 123313.

[9] Cheng D D, Li Y, Xia S Y, et al. A fast granular-ball-based density peaks clustering algorithm for large-scale data[J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35(12): 17202-17215.

[10] Xie J, Kong W Y, Xia S Y, et al. An efficient spectral clustering algorithm based on granular-ball[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(9): 9743-9753.

[11] Sun L, Zhang Q F, Ding W P, et al. FCPFS: fuzzy granular ball clustering-based partial multilabel feature selection with fuzzy mutual information[J]. IEEE Transactions on Emerging Topics in Computational Intelligence, 2025, 9(1): 590-606.

[12] Yang J, Liu Z Z, Xia S Y, et al. 3WC-GBNRS: a novel three-way classifier with granular-ball neighborhood rough sets based on uncertainty[J]. IEEE Transactions on Fuzzy Systems, 2024, 32(8): 4376-4387.

[13] Sun L, Cai Y W, Ding W P, et al. Feature selection-based rough graph classification via graph convolutional neural networks[J]. Engineering Applications of Artificial Intelligence, 2025, 158: 111502.

[14] 丰昌武, 孙林. 融合K-means聚类和标记相关性的多标记Relief特征选择[J]. 聊城大学学报(自然科学版), 2025, 38(1): 122-134.

[15] Feng C W, Sun L. Multilabel relief-based feature selection via fusing K-means clustering and label correlation[J]. Journal of Liaocheng University (Natural Science Edition), 2025, 38(1): 122-134.

[16] Xia S Y, Zhang H, Li W H, et al. GBNRS: a novel rough set algorithm for fast adaptive attribute reduction in classification[J]. IEEE Transactions on Knowledge and Data Engineering, 2022, 34(3): 1231-1242.

[17] Ji X, Peng J H, Zhao P, et al. Extended rough sets model based on fuzzy granular ball and its attribute reduction[J]. Information Sciences, 2023, 640: 119071.

[18] Li Y, Wu X X, Wang X Z . Incremental reduction methods based on granular ball neighborhood rough sets and attribute grouping[J]. International Journal of Approximate Reasoning, 2023, 160: 108974.

[19] Zhang J J, Sun K, Huang B, et al. Attribute reduction based on a rapid variable granular ball generation model[J]. Expert Systems with Applications, 2025, 265: 126030.

[20] Xia S Y, Wang C, Wang G Y, et al. GBRS: a unified granular-ball learning model of pawlak rough set and neighborhood rough set[J]. IEEE Transactions on Neural Networks and Learning Systems, 2025, 36(1): 1719-1733.

[21] Zhang L J, Lin G P, Wei L, et al. A novel ANN-based feature subset selection in multi-scale granular ball neighborhood decision tables[J]. Neural Networks, 2025, 185: 107178.

[22] Yao Y Y . Three-way decisions with probabilistic rough sets[J]. Information Sciences, 2010, 180(3): 341-353.

[23] Deng J, Zhan J M, Herrera-Viedma E, et al. Regret theory-based three-way decision method on incomplete multiscale decision information systems with interval fuzzy numbers[J]. IEEE Transactions on Fuzzy Systems, 2023, 31(3): 982-996.

[24] Savchenko A V, Savchenko L V . Three-way classification for sequences of observations[J]. Information Sciences, 2023, 648: 119540.

[25] Guo D D, Jiang C M, Wu P . Three-way decision based on confidence level change in rough set[J]. International Journal of Approximate Reasoning, 2022, 143: 57-77.

[26] Gao C, Wang Z C, Zhou J, et al. Analysis of core attribute and approximate reduct based on the three-way decision[J]. Applied Soft Computing, 2024, 150: 111117.

基金资助

国家自然科学基金(62506121)

河南省科技攻关项目(262102211015)