Multi-label Feature Selection Based on Fuzzy Neighborhood Similarity Relations in Double Spaces
XU Jiucheng1,2, SHEN Kaili1,2
1. College of Computer and Information Engineering, Henan Nor-mal University, Xinxiang, 453007; 2. Engineering Lab of Intelligence Business and Internet of Things of Henan Province, Henan Normal University, Xinxiang, 453007
Abstract In most of the current rough set based multi-label feature selection algorithms, sample fuzziness and neighborhood relationship are ignored, the neighborhood radius needs setting manually, and attribute importance is measured in a single space. To overcome the defects of classical rough set algorithms, an algorithm of multi-label feature selection based on fuzzy neighborhood similarity in double spaces is proposed from the perspectives of feature space and label space. Firstly, an adaptive neighborhood radius calculation method is proposed and fuzzy neighborhood similarity matrix of samples in feature space is constructed. Secondly, similarities of sample in feature space and label space are obtained according to fuzzy neighborhood similarity relations. Then, the sample similarities in feature space and label space are fused by introducing weights and the attribute importance is calculated based on the fused measures. Finally, a multi-label feature selection algorithm is constructed via the forward greedy algorithm. The effectiveness of the proposed algorithm is confirmed on twelve multi-label datasets.
[1] HUANG S J, GAO W, ZHOU Z H.Fast Multi-instance Multi-label Learning. IEEE Transactions on Pattern Analysis and Machine Inte-lligence, 2019, 41(11): 2614-2627. [2] LIU B S, LIU X L, REN H, et al. Text Multi-label Learning Method Based on Label-Aware Attention and Semantic Dependency. Multimedia Tools and Applications, 2022, 81(5): 7219-7237. [3] 张平照. 多标记学习特征空间和标记空间降维方法研究.硕士学位论文.马鞍山:安徽工业大学, 2020. (ZHANG P Z.Research on Multi-label Learning via Feature Space and Label Space Dimension Reduction Method. Master Dissertation. Maanshan, China: Anhui University of Technology, 2020.) [4] JIANG Z H, LIU K Y, YANG X B, et al. Accelerator for Supervised Neighborhood Based Attribute Reduction. International Journal of Approximate Reasoning, 2020, 119: 122-150. [5] PAWLAK Z. Rough Sets. International Journal of Computer and Information Sciences, 1982, 11(5): 341-356. [6] 段洁,胡清华,张灵均,等.基于邻域粗糙集的多标记分类特征选择算法.计算机研究与发展, 2015, 52(1): 56-65. (DUAN J, HU Q H, ZHANG L J, et al. Feature Selection for Multi-label Classification Based on Neighborhood Rough Sets. Journal Computer Research and Development, 2015, 52(1): 56-65.) [7] 李钰雯. 基于模糊粗糙集模型的特征选择方法研究.博士学位论文.厦门:厦门大学, 2019. (LI Y W.Research on Feature Selection with Fuzzy Rough Sets. Ph.D. Dissertation. Xiamen, China: Xiamen University, 2019.) [8] WANG C Z, SHAO M W, HE Q, et al. Feature Subset Selection Based on Fuzzy Neighborhood Rough Sets. Knowledge-Based Systems, 2016, 111: 173-179. [9] 陈盼盼. 基于粗糙集扩展模型的属性约简算法研究.硕士学位论文.漳州:闽南师范大学, 2020. (CHEN P P.Research on Attribute Reduction Algorithms Based on Extended Rough Set Model. Master Dissertation. Zhangzhou, China: Minnan Normal University, 2020.) [10] SUN L, WANG L Y, DING W P, et al. Feature Selection Using Fuzzy Neighborhood Entropy-Based Uncertainty Measures for Fuzzy Neighborhood Multigranulation Rough Sets. IEEE Transactions on Fuzzy Systems, 2021, 29(1): 19-33. [11] SUN L, WANG L Y, QIAN Y H, et al. Feature Selection Using Lebesgue and Entropy Measures for Incomplete Neighborhood Decision Systems. Knowledge-Based Systems, 2019, 186. DOI: 10.1016/j.knosys.2019.104942. [12] LIN Y J, LI Y W, WANG C X, et al. Attribute Reduction for Multi-label Learning with Fuzzy Rough Set. Knowledge-Based Systems, 2018, 152: 51-61. [13] 赵晋欢,王长忠.基于模糊粗糙集的辨识矩阵属性约简方法.渤海大学学报(自然科学版), 2019, 40(2): 146-151. (ZHAO J H, WANG C Z.Fuzzy Rough Attribute Reduction Method Based on Discernibility Matrix. Journal of Bohai University (Natural Science Edition), 2019, 40(2): 146-151.) [14] 姚二亮,李德玉,李艳红,等.基于双空间模糊辨识关系的多标记特征选择.模式识别与人工智能, 2019, 32(8): 709-717. (YAO E L, LI D Y, LI Y H, et al. Multi-label Feature Selection Based on Fuzzy Discernibility Relations in Double Spaces. Pattern Recognition and Artificial Intelligence, 2019, 32(8): 709-717.) [15] MA J, ATEF M, KHALIL A M, et al. Novel Models of Fuzzy Rough Coverings Based on Fuzzy α-Neighborhood and Its Application to Decision-Making. IEEE Access, 2020, 8: 224354-224364. [16] XU J C, WANG Y, MU H Y, et al. Feature Genes Selection Based on Fuzzy Neighborhood Conditional Entropy. Journal of Intelligent and Fuzzy Systems, 2019, 36(1): 117-126. [17] 吕月姣. 面向多标记数据的邻域自适应粗糙集模型.硕士学位论文.太原:山西大学, 2021. (LÜ Y J.Neighborhood Adaptive Rough Set Model for Multi-label Data. Master Dissertation. Taiyuan, China: Shanxi University, 2021.) [18] ZHANG M L, ZHOU Z H.ML-KNN: A Lazy Learning Approach to Multi-label Learning. Pattern Recognition, 2007, 40(7): 2038-2048. [19] ZHANG Q W, ZHONG Y, ZHANG M L.Feature-Induced Labeling Information Enrichment for Multi-label Learning // Proc of the 32nd AAAI Conference on Artificial Intelligence. Palo Alto, USA: AAAI, 2018: 4446-4453. [20] HUANG R, JIANG W D, SUN G L.Manifold-Based Constraint Laplacian Score for Multi-label Feature Selection. Pattern Recognition Letters, 2018, 112: 346-352. [21] ZHANG Y, ZHOU Z H.Multilabel Dimensionality Reduction via Dependence Maximization. ACM Transactions on Knowledge Discovery from Data, 2010, 4(3). DOI: 10.1145/1839490.1839495. [22] LEE J, KIM D W.Feature Selection for Multi-label Classification Using Multivariate Mutual Information. Pattern Recognition Le-tters, 2013, 34(3): 349-357. [23] SPOLAÔR N, CHERMAN E A, MONARD M C, et al.ReliefF for Multi-label Feature Selection // Proc of the Brazilian Conference on Intelligent Systems. Washington, USA: IEEE, 2013: 6-11. [24] QIAN W B, LONG X D, WANG Y L, et al. Multi-label Feature Selection Based on Label Distribution and Feature Complementarity. Applied Soft Computing, 2020. DOI: 10.1016/j.asoc.2020.106167. [25] XU J C, SHEN K L, SUN L.Multi-label Feature Selection Based on Fuzzy Neighborhood Rough Sets. Complex and Intelligent Systems, 2022, 8: 2105-2129. [26] XU H T, XU L Y.Multi-label Feature Selection Algorithm based on Label Pairwise Ranking Comparison Transformation // Proc of the International Joint Conference on Neural Networks. Washington, USA: IEEE, 2017: 1210-1217. [27] CHANG X J, NIE F P, YANG Y, et al. A Convex Formulation for Semi-supervised Multi-label Feature Selection. Proceedings of the AAAI Conference on Artificial Intelligence, 2014, 28(1): 1171-1177. [28] MA Z G, NIE F P, YANG Y, et al. Web Image Annotation via Subspace-Sparsity Collaborated Feature Selection. IEEE Transactions on Multimedia, 2012, 14(4): 1021-1030. [29] LIM H, LEE J, KIM D W.Optimization Approach for Feature Selection in Multi-label Classification. Pattern Recognition Letters, 2017, 89: 25-30.
2022, Vol. 35 
