Abstract
For a multi-label learning framework, each instance may belong to multiple labels simultaneously. The classification accuracy can be improved significantly by exploiting various correlations, such as label correlations, feature correlations, or the correlations between features and labels. There are few studies on how to combine the feature and label correlations, and they deal more with complete data sets. However, missing labels or other phenomena often occur because of the cost or technical limitations in the data acquisition process. A few label completion algorithms currently suitable for missing multi-label learning, ignore the noise interference of the feature space. At the same time, the threshold of the discriminant function often affects the classification results, especially those of the labels near the threshold. All these factors pose considerable difficulties in dealing with missing labels using label correlations. Therefore, we propose a missing multi-label learning algorithm with non-equilibrium based on a two-level autoencoder. First, label density is introduced to enlarge the classification margin of the label space. Then, a new supplementary label matrix is augmented from the missing label matrix with the non-equilibrium label completion method. Finally, considering feature space noise, a two-level kernel extreme learning machine autoencoder is constructed to implement the information feature and label correlation. The effectiveness of the proposed algorithm is verified by many experiments on both missing and complete label data sets. A statistical analysis of hypothesis validates our approach.
Similar content being viewed by others
References
Gibaja E, Ventura S (2015) A tutorial on multilabel learning[J]. ACM Comput Surv 47(3):1–38
Zhu X, Li X, Zhang S (2016) Block-row sparse multiview multilabel learning for image classification [J]. IEEE Trans Cybern 46(2):450–461
Zhang ML, Zhou ZH (2006) Multilabel neural networks with applications to functional genomics and text categorization [J]. IEEE Trans Knowl Data Eng 18(10):1338–1351
Boutell MR, Luo J, Shen X, Brown CM (2004) Learning multi-label scene classification [J]. Pattern Recogn 37(9):1757–1771
Zhang ML, Zhou ZH (2007) ML-KNN: a lazy learning approach to multi-label learning [J]. Pattern Recogn 40(7):2038–2048
Yu Y, Pedrycz W, Mia D (2014) Multi-label classification by exploiting label correlations [J]. Expert Syst Appl 41(6):2989–3004
Elisseeff A, Weston J (2002) A kernel method for multi-labelled classification [J]. Adv Neural Inf Proces Syst:681–687
Zhang ML, Zhang K (2010) Multi-label learning by exploiting label dependency [C]. In Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 999–1007
Luo FF, Guo WZ, Yu YL, Chen GL (2017) A multi-label classification algorithm based on kernel extreme learning machine [J]. Neurocomputing 260:313–320
Gao N, Huang SJ, Chen S (2016) Multi-label active learning by model guided distribution matching [J]. Front Comput Sci 10(5):845–855
Huang SJ, Chen S, Zhou ZH (2015) Multi-label active learning: Query type matters[C]. In Proceedings of the 24th International Joint Conference on Artificial Intelligence 946–952
Xu M, Jin R, Zhou ZH (2013) Speedup matrix completion with side information: application to multi-label learning [J]. Adv Neural Inf Proces Syst 26:2301–2309
Xu LL, Wang Z, Shen Z, Wang YB, Chen EH (2014) Learning low-rank label correlations for multi-label classification with missing labels [C]. 2014 IEEE International Conference on Data Mining, 1067–1072
Zhu Y, Kwok JT, Zhou ZH (2018) Multi-label learning with global and local label correlation [C]. IEEE Trans Knowl Data Eng 30(6):1081–1094
Huang J, Qin F, Zheng X, Cheng ZK, Yuan ZX, Zhang WG, Huang QM (2019) Improving multi-label classification with missing labels by learning label-specific features [J]. Inf Sci 492:124–146
Cheng YS, Zhao DW, Qian K (2018) Multi-label learning for non-equilibrium labels completion in neighborhood labels space. Pattern Recognit Artif Intell 31(8):740–749
Cheng YS, Zhao DW, Wang YB, Pei GS (2019) Multi-label learning with kernel extreme learning machine autoencoder [J]. Knowl-Based Syst 178:1–10
Cheng YS, Qian K, Wang YB, Zhao DW (2020) Missing multi-label learning with non-equilibrium based on classification margin [J]. Appl Soft Comput 86:105924
Bucak SS, Jin R, Jain AK (2011) Multi-label learning with incomplete class assignments[C]. CVPR 2011. IEEE 2801–2808
Wicker J, Tyukin A, Kramer S (2016) A nonlinear label compression and transformation method for multi-label classification using autoencoders[C], Pacific-Asia Conference on Knowledge Discovery and Data Mining, 328–340
Huang M, Zhuang FZ, Zhang X, Ao X, Niu ZY, Zhang ML, He Q (2019) Supervised representation learning for multi-label classification [J]. Mach Learn 108(5):747–763
Huang GB, Zhu QY, Siew CK (2006) Extreme learning machine: theory and applications [J]. Neurocomputing 70(1–3):489–501
Huang G, Huang GB, Song SJ, You KY (2015) Trends in extreme learning machines: a review [J]. Neural Netw Off J Int Neural Netw Soc 61:32–48
Vincent P, Larochelle H, Bengio Y, Manzagol PA (2008) Extracting and composing robust features with denoising autoencoders [C], proceedings of the 25th international conference on machine learning. ACM, 1096–1103
Zhang N, Ding SF (2017) Unsupervised and semi-supervised extreme learning machine with wavelet kernel for high dimensional data [J]. Memet Comput 8(2):129–139
Zhang N, Ding SF, Shi ZZ (2016) Denoising Laplacian multi-layer extreme learning machine [J]. Neurocomputing 171:1066–1074
Zhang ML, Zhou ZH (2014) A review on multi-label learning algorithms [J]. IEEE Trans Knowl Data Eng 26(8):1819–1837
Ding SF, Zhang N, Zhang J, Xu XZ, Shi ZZ (2017) Unsupervised extreme learning machine with representational features [J]. Int J Mach Learn Cybern 8(2):587–595
Ding SF, Guo LL, Hou YL (2017) Extreme learning machine with kernel model based on deep learning[J]. Neural Comput & Applic 28(8):1975–1984
Deng WY, Zheng QH, Chen L, Xu XB (2010) Research on extreme learning of neural networks [J]. Chin J Comput 33(2):279–287
Chen NY, Liu Y, Chen HQ, Cheng JJ (2017) Detecting communities in social networks using label propagation with information entropy [J]. Phys A Stat Mech Appl 471:788–798
Li DD, Wang Z, Cao C, Liu Y (2018) Information entropy based sample reduction for support vector data description [J]. Appl Soft Comput 71:1153–1160
Lu X, Tsao Y, Matsuda S, Hori C (2013) Speech enhancement based on deep denoising autoencoder [C], Interspeech, 436–440
Tsoumakas G, Katakis L, Vlahavas L (2011) Random k-Labelsets for multilabel classification [J]. IEEE Trans Knowl Data Eng 23(7):1079–1089
Zhang ML, Pena JM, Robles V (2009) Feature selection for multi-label naive Bayes classification [J]. Inf Sci 179(19):3218–3229
Zhang ML, Wu L (2015) Lift: multi-label learning with label-specific features [J]. IEEE Trans Pattern Anal Mach Intell 37(1):107–120
Demsar J (2006) Statistical comparisons of classifiers over multiple data sets [J]. J Mach Learn Res 7(1):1–30
Williamson DF, Parker RA, Kendrick JS (1989) The box plot: a simple visual method to interpret data [J]. Ann Intern Med 110(11):916–921
Acknowledgements
This research is supported by Key Laboratory of Intelligent Computing & Signal Processing, Ministry of Education (Anhui University)(2020A003).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Cheng, Y., Song, F. & Qian, K. Missing multi-label learning with non-equilibrium based on two-level autoencoder. Appl Intell 51, 6997–7015 (2021). https://doi.org/10.1007/s10489-020-02140-1
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10489-020-02140-1