基于全局和局部注意力的多维分类Multi-Dimensional Classification Based on Global and Local Attention
程玉胜,王佳宝
摘要(Abstract)：
在多维分类中，单个实例同时与多个类变量相关联，且每一个类变量来自不同维度。文章引入注意力机制，同时利用类变量和实例信息，以提高多维分类任务的质量。首先，使用一一分解技术以对齐类变量空间；其次，使用注意力机制来对分解后的类变量空间和实例信息进行编码，并在编码后的空间上进行注意力计算，以得到信息增强后的类变量空间；最后，利用类变量相关性的全局和局部特征以提高模型的鲁棒性，并完成分类任务。在多个基准多维数据集上的实验分析表明，该方法相对于现有算法具有一定的优势。
关键词(KeyWords)： 机器学习;多维分类;注意力机制;标签编码
基金项目(Foundation): 安徽省自然科学基金项目（2108085MF216）
作者(Author): 程玉胜,王佳宝
DOI: 10.13757/j.cnki.cn34-1328/n.2025.01.013
参考文献(References)：
- [1] XU D, SHI Y, TSANG I W, et al. Survey on multi-output learning[J]. IEEE Transactions on Neural Networks and Learning Systems, 2019,31(7):2409-2429.
- [2]陈飞,史金成.粗糙互信息的不平衡多标记特征选择算法[J].安庆师范大学学报(自然科学版), 2021, 27(1):40-43, 58.
- [3]程玉胜,张露露,王一宾,等.特征特定标记关联挖掘的类属属性学习[J].计算机研究与发展, 2021, 58(1):34-47.
- [4] JIA B B, ZHANG M L. Multi-dimensional classification via kNN feature augmentation[J]. Pattern Recognition, 2020, 106:107423.
- [5] SERAFINO F, PIO G, CECI M, et al. Hierarchical multidimensional classification of web documents with multiwebclass[C]. Discovery Science:18th International Conference, 2015.
- [6] MUKTADIR A H A, MIYAZAWA T, MARTINEZ-JULIA P, et al. Multi-target classification based automatic virtual resource allocation scheme[J]. IEICE Transactions on Information and Systems, 2019, 102(5):898-909.
- [7] JIA B B, ZHANG M L. MD-KNN:an instance-based approach for multi-dimensional classification[C]. 2020 25th International Conference on Pattern Recognition, IEEE, 2021.
- [8] GIL-BEGUE S, BIELZA C, LARRA?AGA P. Multi-dimensional Bayesian network classifiers:a survey[J]. Artificial Intelligence Review,2021, 54(1):519-559.
- [9] READ J, BIELZA C, LARRA?AGA P. Multi-dimensional classification with super-classes[J]. IEEE Transactions on Knowledge and Data Engineering, 2013, 26(7):1720-1733.
- [10]高广尚.推荐系统中神经网络结合注意力机制研究综述[J].计算机工程与应用, 2024, 60(10):47-60.
- [11] CHENG Y S, QIAN K, MIN F. Global and local attention-based multi-label learning with missing labels[J]. Information Sciences, 2022,594:20-42.
- [12] JIA B B, ZHANG M L. Multi-dimensional classification via decomposed label encoding[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(2):1844-1856.
- [13] ZHU Y, KWOK J T, ZHOU Z H. Multi-label learning with global and local label correlation[J]. IEEE Transactions on Knowledge and Data Engineering, 2017, 30(6):1081-1094.
- [14] BOUMAL N, MISHRA B, ABSIL P A, et al. Manopt, a Matlab toolbox for optimization on manifolds[J]. The Journal of Machine Learning Research, 2014, 15(1):1455-1459.
- [15] JIA B B, ZHANG M L. Multi-dimensional classification via stacked dependency exploitation[J]. Science China Information Sciences,2020, 63:1-14.
- [16] READ J, MARTINO L, LUENGO D. Efficient monte carlo methods for multi-dimensional learning with classifier chains[J]. Pattern Recognition, 2014, 47(3):1535-1546.
- [17] DEM?AR J. Statistical comparisons of classifiers over multiple data sets[J]. The Journal of Machine Learning Research, 2006, 7:1-30.