RJXB软件学报Journal of Software1000-9825软件学报编辑部中国北京rjxb-34-4-173210.13328/j.cnki.jos.006703TP18模式识别与人工智能Pattern Recognition and Artificial Intelligence利用标签相关性先验的弱监督多标签学习方法Weakly Supervised Multi-label Learning Using Prior Label Correlation Information欧阳宵OUYANGXiao
Multi-label learning is a very important machine learning paradigm. Traditional multi-label learning methods are designed in supervised or semi-supervised manner. Generally, they require accurate labeling of all or partial data into multiple categories. In many practical applications, it is difficult to obtain the label information with a large number of labels, which greatly restricts the promotion and application of multi-label learning. In contrast, label correlation, as a common weak supervision information, has lower requirements for labeling information. How to use label correlation for multi-label learning is an important but unstudied problem. This study proposes a method named weakly supervised multi-label learning using prior label correlation information (WSMLLC). This model restates the sample similarity by using label correlation, and can obtain label indicator matrix effectively, constrain the projection matrix of data by using prior information, and modify the indicator matrix by introducing regression terms. Compared with the existing methods, the outstanding advantage of WSMLLC model is that it can realize the label assignment of multi-label samples only by providing label correlation priors. Experimental results show that WSMLLC has obvious advantages over current advanced multi-label learning methods in the case of complete loss of label matrix.
ReferencesZhang ML, Zhou ZHA review on multi-label learning algorithms20142681819183710.1109/TKDE.2013.39
Zhang ML, Zhou ZH. A review on multi-label learning algorithms. IEEE Trans. on Knowledge Data Engineering, 2014, 26(8): 1819−1837. [doi: 10.1109/TKDE.2013.39]
Elisseeff A, Weston JA kernel method for multi-labelled classification200114681687
Elisseeff A, Weston J. A kernel method for multi-labelled classification. Advances in Neural Information Processing Systems, 2001, 14: 681−687.
Xiao L, Chen BL, Huang X, et alMulti-label text classification method based on label semantic information20203141079108910.13328/j.cnki.jos.005923
Xiao L, Chen BL, Huang X, et al. Multi-label text classification method based on label semantic information. Ruan Jian Xue Bao/Journal of Software, 2020, 31(4): 1079−1089 (in Chinese with English abstract). http://www.jos.org.cn/1000-9825/5923.htm [doi: 10.13328/j.cnki.jos.005923]
Zhang JJ, Wu Q, Shen CH, et alMulti-Label image classification with regional latent semantic dependencies201820102801281310.1109/TMM.2018.2812605
Zhang JJ, Wu Q, Shen CH, et al. Multi-Label image classification with regional latent semantic dependencies. IEEE Trans. on Multimedia, 2018, 20(10): 2801−2813. [doi: 10.1109/TMM.2018.2812605]
Wang Y, Wu YJ, Wu JZ, et alMulti-dimensional tag recommender model via heterogeneous networks201728102611262410.13328/j.cnki.jos.005132
Wang Y, Wu YJ, Wu JZ, et al. Multi-dimensional tag recommender model via heterogeneous networks. Ruan Jian Xue Bao/Journal of Software, 2017, 28(10): 2611−2624 (in Chinese with English abstract). http://www.jos.org.cn/1000-9825/5132.htm [doi: 10.13328/j.cnki.jos.005132]
Shan JC, Hou CP, Zhuge WZ, et alCo-learning binary classifiers for LP-based multi-label classification20195514615210.1016/j.cogsys.2019.01.009
Shan JC, Hou CP, Zhuge WZ, et al. Co-learning binary classifiers for LP-based multi-label classification. Cognitive Systems Research, 2019, 55: 146−152.
Tsoumakas G, Vlahavas IP. Random k-labelsets: An ensemble method for multilabel classification. In: Proc. of the 18th European Conf. on Machine Learning, Vol. 4701. Warsaw: Springer, 2007. 406−417.
Zhang ML, Zhou ZHML-KNN: A lazy learning approach to multi-label learning20074072038204810.1016/j.patcog.2006.12.019
Zhang ML, Zhou ZH. ML-KNN: A lazy learning approach to multi-label learning. Pattern Recognition, 2007, 40(7): 2038-2048.
Liu Y, Jin R, Yang LSemi-supervised multi-label learning by constrained non-negative matrix factorization20066421426
Liu Y, Jin R, Yang L. Semi-supervised multi-label learning by constrained non-negative matrix factorization. Association for the Advance of Artificial Intelligence, 2006, 6: 421−426.
Huang J, Xu L, Qian K, et alMulti-label learning with missing and completely unobserved labels20213531061108610.1007/s10618-021-00743-x
Huang J, Xu L, Qian K, et al. Multi-label learning with missing and completely unobserved labels. Data Mining and Knowledge Discovery, 2021, 35(3): 1061−1086.
Luo FF, Guo WZ, Chen GLAddressing imbalance in weakly supervised multi-label learning20197374633747210.1109/ACCESS.2019.2906409
Zhang ML, Li YK, Liu XY, et alBinary relevance for multi-label learning: An overview201812219120210.1007/s11704-017-7031-7
Zhang ML, Li YK, Liu XY, et al. Binary relevance for multi-label learning: An overview. Frontiers of Computer Science, 2018, 12(2): 191−202.
Weng W, Lin Y, Wu S, et alMulti-label learning based on label-specific features and local pairwise label correlation201827338539410.1016/j.neucom.2017.07.044
Weng W, Lin Y, Wu S, et al. Multi-label learning based on label-specific features and local pairwise label correlation. Neurocomputing, 2018, 273: 385−394.
Zhao W, Kong S, Bai J, et alHOT-VAE: Learning high-order label correlation for multi-label classification via attention-based variational autoencoders20211501615024
Zhao W, Kong S, Bai J, et al. HOT-VAE: Learning high-order label correlation for multi-label classification via attention-based variational autoencoders. Association for the Advance of Artificial Intelligence, 2021, 15016−15024.
Nodet P, Lemaire V, Bondu A, et al. From weakly supervised learning to biquality learning, an introduction. In: Proc. of the 2021 Int'l Joint Conf. on Neural Networks. Shenzhen: IEEE, 2021. 1−10. [doi: 10.1109/IJCNN52387.2021.9533353]
Frénay B, Verleysen MClassification in the presence of label noise: A survey201425584586910.1109/TNNLS.2013.2292894
Frénay B, Verleysen M. Classification in the presence of label noise: A survey. IEEE Trans. on Neural Networks and Learning Systems, 2014, 25(5): 845−869. [doi: 10.1109/TNNLS.2013.2292894]
Sun LJ, Lyu G, Feng SH, et alBeyond missing: Weakly-supervised multi-label learning with incomplete and noisy labels20215131552156410.1007/s10489-020-01878-y
Sun LJ, Lyu G, Feng SH, et al. Beyond missing: Weakly-supervised multi-label learning with incomplete and noisy labels. Applied Intelligence, 2021, 51(3): 1552−1564.
Guo HF, Han L, Su S, et alDeep multi-instance multi-label learning for image annotation201832318590011859005
Guo HF, Han L, Su S, et al. Deep multi-instance multi-label learning for image annotation. Int'l Journal of Pattern Recognition and Artificial Intelligence, 2018, 32(3): 1859001−1859005.
Zeng Z, Gao N, Xue C, et al. Learning from audience interaction: Multi-instance multi-label topic model for video shots annotating. In: Proc. of the 24th IEEE Int'l Conf. on Computer Supported Cooperative Work in Design. Dalian: IEEE, 2021. 1075−1080. [doi: 10.1109/CSCWD49262.2021.9437805]
Lv G, Xu T, Chen E, et al. Reading the videos: Temporal labeling for crowdsourced time-sync videos based on semantic Embedding. In: Proc. of the AAAI Conf. on Artificial Intelligence. 2016, 30(1): 3000−3006.
Xu KX, Zhao ZY, Gu JP, et al. Multi-instance multi-label learning for gene mutation prediction in hepatocellular carcinoma. In: Proc. of the 42nd Annual Int'l Conf. of the IEEE Engineering in Medicine & Biology Society. 2020. 6095−6098. [doi: 10.1109/EMBC44109.2020.9175293]
Huang J, Xu L, Wang J, et al. Discovering latent class labels for multi-label learning. In: Proc. of the Int'l Joint Conf. on Artificial Intelligence Organization. 2020. 3058−3064. [doi: https://doi.org/10.24963/ijcai.2020/423
]
Cheng Z, Zeng ZJoint label-specific features and label correlation for multi-label learning with missing label202050114029404910.1007/s10489-020-01715-2
Cheng Z, Zeng Z. Joint label-specific features and label correlation for multi-label learning with missing label. Applied Intelligence, 2020, 50(11): 4029−4049.
Zhu Y, Kwok JT, Zhou ZMulti-label learning with global and local label correlation20173061081109410.1109/TKDE.2017.2785795
Zhu Y, Kwok JT, Zhou Z. Multi-label learning with global and local label correlation. IEEE Trans. on Knowledge and Data Engineering, 2017, 30(6): 1081−1094. [doi: 10.1109/TKDE.2017.2785795]
Mao K, Niu J, Liu X, et al. Word2Cluster: A new multi-label text clustering algorithm with an adaptive clusters number. In: Proc. of the 2019 IEEE Global Communications Conf. Waikoloa: IEEE, 2019. 1−6. [doi: 10.1109/GLOBECOM38437.2019.9013266]
Boyd S, Vandenberghe L. Convex Optimization. Cambridge University Press, 2004.
Fan RD, Luo TJ, Zhuge WZ, et alMulti-view subspace learning via bidirectional sparsity202010810752410.1016/j.patcog.2020.107524
Fan RD, Luo TJ, Zhuge WZ, et al. Multi-view subspace learning via bidirectional sparsity. Pattern Recognition, 2020, 108: 107524.
Zhang ML, Fang JP, Wang YBBiLabel-specific features for multi-label classification202218123
Zhang ML, Fang JP, Wang YB. BiLabel-specific features for multi-label classification. ACM Trans. on Knowledge Discovery from Data, 2022, 18: 1−23.
Zhang ML, Fang JPPartial multi-label learning via credible label elicitation202143103587359910.1109/TPAMI.2020.2985210
Zhang ML, Fang JP. Partial multi-label learning via credible label elicitation. IEEE Trans. on Pattern Analysis and Machine Intelligence, 2021, 43(10): 3587−3599. [doi: 10.1109/TPAMI.2020.2985210]
Li SN, Li N, Li ZHMulti-label data mining: A survey20134041421
Li SN, Li N, Li ZH. Multi-label data mining: A survey. Computer Science, 2013, 40(4): 14−21 (in Chinese with English abstract).
Sun N, Shan J, Hou C. Multi-label active learning with error correcting output codes. In: Proc. of the Pacific-Asia Conf. on Knowledge Discovery and Data Mining. 2019. 331−342.