Abstract
Class hierarchical structures play a significant role in large and complex tasks of machine learning. Existing studies on the construction of such structures follow a two-stage strategy. The category similarities are first computed with a certain assumption, and the group partition algorithm is then performed with some hyper-parameters to control the shape of class hierarchy. Despite their effectiveness in many cases, these methods suffer from two problems: (1) optimizing the two-stage objective to obtain the structure is sub-optimal; (2) hyper-parameters make the search space too large to find the optimal structure efficiently. In this paper, we propose a unified and dynamic framework to address these problems, which can: (1) jointly optimize the category similarity and group partition; (2) obtain the class hierarchical structure dynamically without any hyper-parameters. The framework replaces the traditional category similarity with the sample similarity, and constrains samples from the same atomic category partitioned to the same super-category. We theoretically prove that, within our framework, the sample similarity is equivalent to the category similarity and can balance the partitions in terms of the number of samples. Further, we design a modularity-based partition optimization algorithm that can automatically determine the number of partitions on each level. Extensive experimental results on multiple image classification datasets show that the hierarchical structure constructed by the proposed method achieves better accuracy and efficiency compared to existing methods. Additionally, the hierarchy obtained by the proposed method can benefit long-tail learning scenarios due to the balanced partition on samples.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.Data availability
All data used in this paper is publicly available and can be found in the cited paper.
Code availability
Code is available at https://github.com/wangyuTJU/greedyIsolation.
References
Zhai J, Zhang S, Wang C (2017) The classification of imbalanced large data sets based on mapreduce and ensemble of elm classifiers. Int J Mach Learn Cybern 8(3):1009–1017
Dabbu M, Karuppusamy L, Pulugu D, Vootla SR, Reddyvari VR (2022) Water atom search algorithm-based deep recurrent neural network for the big data classification based on spark architecture. Int J Mach Learn Cybern 13(8):2297–2312
Pan L, Wang S, Ding Y, Zhao L, Song A (2022) A universal emotion recognition method based on feature priority evaluation and classifier reinforcement. Int J Mach Learn Cybern 13(10):3225–3237
Zheng Y, Fan J, Zhang J, Gao X (2017) Hierarchical learning of multi-task sparse metrics for large-scale image classification. Pattern Recogn 67:97–109
Zhou Y, Hu Q, Wang Y (2018) Deep super-class learning for long-tail distributed image classification. Pattern Recogn 80:118–128
Lin Y, Liu H, Zhao H, Hu Q, Zhu X, Wu X (2022) Hierarchical feature selection based on label distribution learning. IEEE Transact Knowledge Data Eng
Han S, Mao H, Dally WJ (2016) Deep compression: compressing deep neural networks with pruning, trained quantization and huffman coding. In: ICLR 2016 : International Conference on Learning Representations 2016
Deng J, Krause J, Berg AC, Fei-Fei L (2012) Hedging your bets: Optimizing accuracy-specificity trade-offs in large scale visual recognition. In: IEEE Conference on Computer Vision and Pattern Recognition, pp. 3450–3457. https://doi.org/10.1109/CVPR.2012.6248086
Tenenbaum JB, Kemp C, Griffiths TL, Goodman ND (2011) How to grow a mind: statistics, structure, and abstraction. Science 331(6022):1279–1285
Lin Y, Hu Q, Liu J, Zhu X, Wu X (2021) Mulfe: multi-label learning via label-specific feature space ensemble. ACM Transact Knowledge Discovery Data (TKDD) 16(1):1–24
Bellmund JL, Gärdenfors P, Moser EI, Doeller CF (2018) Navigating cognition: spatial codes for human thinking. Science 362(6415):6766
Ye Q, Shi W, Qu K, He H, Zhuang W, Shen X (2021) Joint ran slicing and computation offloading for autonomous vehicular networks: a learning-assisted hierarchical approach. IEEE Open J Vehicular Technol 2:272–288
Al-taezi M, Zhu P, Hu Q, Wang Y, Al-Badwi A (2021) Self-paced hierarchical metric learning (sphml). Int J Mach Learn Cybern 12(9):2529–2541
Xu Z, Zhang B, Li D, Yue X (2022) Hierarchical multilabel classification by exploiting label correlations. Int J Mach Learn Cybern 13(1):115–131
Fu S, Wang G, Xu J (2021) hier2vec: interpretable multi-granular representation learning for hierarchy in social networks. Int J Mach Learn Cybern 12(9):2543–2557
Zhang X, Zhou Y, Tang X, Fan Y (2022) Three-way improved neighborhood entropies based on three-level granular structures. Int J Mach Learn Cybern 13(7):1861–1890
Deng J, Dong W, Socher R, Li L-J, Li K, Fei-Fei L (2009) Imagenet: a large-scale hierarchical image database. In: 2009 IEEE Conference on Computer Vision and Pattern Recognition, pp. 248–255
Chen H, Wang Y, Hu Q (2022) Multi-granularity regularized re-balancing for class incremental learning. IEEE Transact Knowledge Data Eng. https://doi.org/10.1109/TKDE.2022.3188335
Wang Y, Wang Z, Hu Q, Zhou Y, Su H (2022) Hierarchical semantic risk minimization for large-scale classification. IEEE Transact Cybern 52(9):9546–9558. https://doi.org/10.1109/TCYB.2021.3059631
Bengio S, Weston J, Grangier D (2010) Label embedding trees for large multi-class tasks. In: Advances in Neural Information Processing Systems 23, pp. 163–171
Liu Y, Dou Y, Jin R, Li R (2018) Visual confusion label tree for image classification. In: 2018 IEEE International Conference on Multimedia and Expo (ICME), pp. 1–6
Zhao H, Guo S, Lin Y (2021) Hierarchical classification of data with long-tailed distributions via global and local granulation. Inf Sci 581:536–552. https://doi.org/10.1016/j.ins.2021.09.059
Fellbaum C (2000) Wordnet : an electronic lexical database. Language 76(3):706
Zhang C, Cheng J, Tian Q (2018) Image-level classification by hierarchical structure learning with visual and semantic similarities. Inf Sci 422:271–281. https://doi.org/10.1016/j.ins.2017.09.024
Li L-J, Wang C, Lim Y, Blei DM, Fei-Fei L (2010) Building and using a semantivisual image hierarchy. In: 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 3336–3343
Naphade M, Smith JR, Tesic J, Chang S-F, Hsu W, Kennedy L, Hauptmann A, Curtis J (2006) Large-scale concept ontology for multimedia. IEEE Multimedia 13(3):86–91
Sun M, Huang W, Savarese S (2013) Find the best path: An efficient and accurate classifier for image hierarchies. In: 2013 IEEE International Conference on Computer Vision, pp. 265–272
Lei H, Mei K, Zheng N, Dong P, Zhou N, Fan J (2014) Learning group-based dictionaries for discriminative image representation. Pattern Recogn 47(2):899–913
Griffin G, Perona P (2008) Learning and using taxonomies for fast visual categorization. In: 2008 IEEE Conference on Computer Vision and Pattern Recognition, pp. 1–8
Yan Z, Zhang H, Piramuthu R, Jagadeesh V, DeCoste D, Di W, Yu Y (2015) Hd-cnn: Hierarchical deep convolutional neural networks for large scale visual recognition. In: 2015 IEEE International Conference on Computer Vision (ICCV), pp. 2740–2748
Deng J, Satheesh S, Berg AC, Li F (2011) Fast and balanced: Efficient label tree learning for large scale object recognition. In: Advances in Neural Information Processing Systems 24, pp. 567–575
Liu B, Sadeghi F, Tappen M, Shamir O, Liu C (2013) Probabilistic label trees for efficient large scale image classification. In: 2013 IEEE Conference on Computer Vision and Pattern Recognition, pp. 843–850
Fan J, Zhou N, Peng J, Gao L (2015) Hierarchical learning of tree classifiers for large-scale plant species identification. IEEE Trans Image Process 24(11):4172–4184
Fan J, Zhao T, Kuang Z, Zheng Y, Zhang J, Yu J, Peng J (2017) Hd-mtl: hierarchical deep multi-task learning for large-scale visual recognition. IEEE Trans Image Process 26(4):1923–1938
Qu Y, Lin L, Shen F, Lu C, Wu Y, Xie Y, Tao D (2017) Joint hierarchical category structure learning and large-scale image classification. IEEE Trans Image Process 26(9):4331–4346
Frey BJ, Dueck D (2007) Clustering by passing messages between data points. Science 315(5814):972–976
Zheng Y, Chen Q, Fan J, Gao X (2020) Hierarchical convolutional neural network via hierarchical cluster validity based visual tree learning. Neurocomputing 409:408–419
Blondel VD, Guillaume JL, Lambiotte R, Lefebvre E (2008) Fast unfolding of communities in large networks. J Stat Mech 2008(10):10008
Newman MEJ (2004) Fast algorithm for detecting community structure in networks. Phys Rev E 69(6):066133
Clauset A, Newman MEJ, Moore C (2004) Finding community structure in very large networks. Phys Rev E 70(6):66111
Brandes U, Delling D, Gaertler M, Goerke R, Hoefer M, Nikoloski Z, Wagner D (2006) Maximizing modularity is hard. arXiv preprint arXiv:physics/0608255
Wang S, Siskind JM (2003) Image segmentation with ratio cut. IEEE Trans Pattern Anal Mach Intell 25(6):675–690
Krizhevsky A (2009) Learning Multiple Layers of Features from Tiny Images. Master thesis
Xiao J, Hays J, Ehinger KA, Oliva A, Torralba A (2010) Sun database: Large-scale scene recognition from abbey to zoo. In: 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 3485–3492
Pouransari H, Ghili S (2014) Tiny imagenet visual recognition challenge. CS 231N
Cui Y, Jia M, Lin T-Y, Song Y, Belongie S (2019) Class-balanced loss based on effective number of samples. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 9268–9277
He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778
Clauset A, Moore C, Newman MEJ (2008) Hierarchical structure and the prediction of missing links in networks. Nature 453(7191):98–101
Wang Y, Liu R, Lin D, Chen D, Li P, Hu Q, Chen CLP (2021) Coarse-to-fine: Progressive knowledge transfer-based multitask convolutional neural network for intelligent large-scale fault diagnosis. IEEE Transactions on Neural Networks and Learning Systems, 1–14. https://doi.org/10.1109/TNNLS.2021.3100928
Wu A, Han Y, Zhu L, Yang Y (2021) Instance-invariant domain adaptive object detection via progressive disentanglement. IEEE Trans Pattern Anal Mach Intell 44(8):4178–4193
Wu A, Zhao S, Deng C, Liu W (2021) Generalized and discriminative few-shot object detection via svd-dictionary enhancement. Adv Neural Inf Process Syst 34:6353–6364
Acknowledgements
This work was supported in part by the National Natural Scientific Foundation of China (NSFC) under Grants 62106174, and 61732011, and in part by the China Postdoctoral Science Foundation under Grants 2021TQ0242 and 2021M690118.
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
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Huang, H., Wang, Y. & Hu, Q. Building hierarchical class structures for extreme multi-class learning. Int. J. Mach. Learn. & Cyber. 14, 2575–2590 (2023). https://doi.org/10.1007/s13042-023-01783-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s13042-023-01783-z