基于联合信息保持的异构领域自适应

doi:10.3778/j.issn.1673-9418.1905085

摘要/Abstract

摘要：

异构领域自适应（HDA）的主要目的是借助源域的知识，辅助处于不同特征空间中目标域的数据进行建模。异构领域自适应一个核心的问题是如何在领域适配过程中有效保持原始数据的信息，减少因为特征变换导致的信息损失，提出了一个联合信息保持算法（JIP）来解决上述问题。所提算法通过两方面来保持原始数据的信息：一方面，在异构领域自适应场景中，两个领域之间通常存在大量配对样本，所提算法通过最大化配对样本之间的相关性来保持这种配对信息。另一方面，所提算法优化了结构信息保持策略，同时保持了原始数据的局部结构信息和全局结构信息。最终，将联合信息保持和分布匹配整合在一起，从而实现异构领域自适应。实验结果表明，所提算法较之于先进的异构领域自适应算法有明显优势。

关键词: 异构领域自适应（HDA）, 信息损失, 配对信息

Abstract:

Heterogeneous domain adaptation (HDA) aims to assist the modeling tasks of the target domain lied in different spaces with knowledge of the source domain. A core issue of HDA is how to preserve the information of the original data during domain adaptation and eliminate the information loss caused by feature transformation. This paper proposes a joint information preservation (JIP) method to deal with the problem. The method preserves the information of the original data from two aspects. On the one hand, large number of paired samples often exist between the two domains of the HDA and the proposed method preserves the paired information by maximizing the correlation of the paired samples. On the other hand, the proposed method improves the strategy of preserving the structural information of the original data, where the local and global structural information are preserved simultaneously. Finally, the JIP is integrated with distribution matching to achieve HDA tasks. Experimental results show the superiority of the proposed method over the state-of-the-art HDA algorithms.

Key words: heterogeneous domain adaptation (HDA), information loss, paired information

许鹏，邓赵红，王骏，王士同. 基于联合信息保持的异构领域自适应[J]. 计算机科学与探索, 2020, 14(7): 1183-1193.

XU Peng, DENG Zhaohong, WANG Jun, WANG Shitong. Joint Information Preservation for Heterogeneous Domain Adaptation[J]. Journal of Frontiers of Computer Science and Technology, 2020, 14(7): 1183-1193.

参考文献

[1] Csurka G. Domain adaptation for visual applications: a com-prehensive survey[J]. arXiv:1702.05374, 2017.
[2] Weiss K R, Khoshgoftaar T M, Wang D D. A survey of trans-fer learning[J]. Journal of Big Data, 2016, 3: 9.
[3] Day O, Khoshgoftaar T M. A survey on heterogeneous trans-fer learning[J]. Journal of Big Data, 2017, 4: 29.
[4] Moon S, Carbonell J G. Completely heterogeneous transfer learning with attention: what and what not to transfer[C]//Proceedings of the 26th International Joint Conference on Artificial Intelligence, Melbourne, Aug 19-25, 2017: 2508-2514.
[5] Long M S, Wang J M, Ding G G, et al. Transfer joint matc-hing for unsupervised domain adaptation[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition, Columbus, Jun 21-23, 2014. Washington: IEEE Computer Society, 2014: 1410-1417.
[6] Aljundi R, Emonet R, Muselet D, et al. Landmarks-based kernelized subspace alignment for unsupervised domain ada-ptation[C]//Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition, Boston, Jun 7-12, 2015. Washington: IEEE Computer Society, 2015: 56-63.
[7] Long M S, Wang J M, Ding G G, et al. Adaptation regula-rization: a general framework for transfer learning[J]. IEEE Transactions on Knowledge and Data Engineering, 2014, 26(5): 1076-1089.
[8] Wang J D, Feng W J, Chen Y Q, et al. Visual domain adap-tation with manifold embedded distribution alignment[J].arXiv:1807.07258, 2018.
[9] Sun B C, Feng J S, Saenko K. Return of frustratingly easy domain adaptation[C]//Proceedings of the 30th AAAI Con-ference on Artificial Intelligence, Phoenix, Feb 12-17, 2016. Menlo Park: AAAI, 2016: 2058-2065.
[10] Yan K, Kou L, Zhang D. Learning domain-invariant subspace using domain features and independence maximization[J]. IEEE Transactions on Systems, Man, and Cybernetics, 2018, 48(1): 288-299.
[11] Zhong E H, Fan W, Peng J, et al. Cross domain distribution adaptation via kernel mapping[C]//Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Dis-covery and Data Mining, Paris, Jun 28 - Jul 1, 2009. New York: ACM, 2009: 1027-1036.
[12] Fernando B, Habrard A, Sebban M, et al. Subspace alignment for domain adaptation[J]. arXiv:1409.5241, 2014.
[13] Friedjungová M, Ji?ina M. Asymmetric heterogeneous trans-fer learning: a survey[C]//Proceedings of the 6th International Conference on Data Science, Technology and Applications, Madrid, Jul 24-26, 2017: 17-27.
[14] Xiao M, Guo Y H. Feature space independent semi-supervised domain adaptation via kernel matching[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(1): 54-66.
[15] Duan L X, Tsang I W, Xu D. Domain transfer multiple ker-nel learning[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(3): 465-479.
[16] Chen Y M, Song S J, Li S, et al. Domain space transfer extr-eme learning machine for domain adaptation[J]. IEEE Trans-actions on Systems, Man, and Cybernetics, 2019, 49(5): 1909-1922.
[17] Deng Z H, Xu P, Xie L X, et al. Transductive joint know-ledge transfer TSK FS for recognition of epileptic EEG sig-nals[J]. IEEE Transactions on Neural Systems and Rehabilita-tion Engineering, 2018, 26(8): 1481-1494.
[18] Li J J, Lu K, Huang Z, et al. Transfer independently together: a generalized framework for domain adaptation[J]. IEEE Transactions on Cybernetics, 2019, 49(6): 2144-2155.
[19] Si S, Tao D C, Geng B. Bregman divergence-based regulari-zation for transfer subspace learning[J]. IEEE Transactions on Knowledge and Data Engineering, 2010, 22(7): 929-942.
[20] Zhuang F Z, Cheng X H, Luo P, et al. Supervised represen-tation learning: transfer learning with deep autoencoders[C]// Proceedings of the 24th International Joint Conference on Artificial Intelligence, Buenos Aires, Jul 25-31, 2015. Menlo Park: AAAI, 2015: 4119-4125.
[21] Shen J, Qu Y R, Zhang W N, et al. Wasserstein distance gui-ded representation learning for domain adaptation[C]//Procee-dings of the 32nd AAAI Conference on Artificial Intelligence, the 30th Innovative Applications of Artificial Intelligence,and the 8th AAAI Symposium on Educational Advances in Artificial Intelligence, New Orleans, Feb 2-7, 2018. Menlo Park: AAAI, 2018: 4058-4065.
[22] Tahmoresnezhad J, Hashemi S. Visual domain adaptation via transfer feature learning[J]. Knowledge and Information Systems, 2017, 50(2): 585-605.
[23] Wang C, Mahadevan S. Heterogeneous domain adaptation using manifold alignment[C]//Proceedings of the 22nd Inter-national Joint Conference on Artificial Intelligence, Barce-lona, Jul 16-22, 2011. Menlo Park: AAAI, 2011: 1541-1546.
[24] Yeh Y R, Huang C H, Wang Y C F. Heterogeneous domain adaptation and classification by exploiting the correlation subspace[J]. IEEE Transactions on Image Processing, 2014, 23(5): 2009-2018.
[25] Yan Y G, Li W, Ng M K P, et al. Learning discriminative correlation subspace for heterogeneous domain adaptation [C]//Proceedings of the 26th International Joint Conference on Artificial Intelligence, Melbourne, Aug 19-25, 2017: 3252-3258.
[26] Mehrkanoon S, Suykens J A K. Regularized semipaired ker-nel CCA for domain adaptation[J]. IEEE Transactions on Neu-ral Networks and Learning Systems, 2018, 29: 3199-3213.
[27] Shi X, Liu Q, Fan W, et al. Transfer learning on heterogene-ous feature spaces via spectral transformation[C]//Proceed-ings of the 10th IEEE International Conference on Data Min-ing, Sydney, Dec 14-17, 2010: 1049-1054.
[28] Kulis B, Saenko K, Darrell T. What you saw is not what you get: domain adaptation using asymmetric kernel trans-forms[C]//Proceedings of the 24th IEEE Conference on Computer Vision and Pattern Recognition, Colorado Springs, Jun 20-25, 2011. Washington: IEEE Computer Society, 2011: 1785-1792.
[29] Hoffman J, Rodner E, Donahue J, et al. Efficient learning of domain-invariant image representations[C]//Proceedings of the 1st International Conference on Learning Representations, Scottsdale, May 2-4, 2013: 1-42.
[30] Li W, Duan L X, Xu D, et al. Learning with augmented features for supervised and semi-supervised heterogeneous domain adaptation[J]. IEEE Transactions on Pattern Analy-sis and Machine Intelligence, 2014, 36(6): 1134-1148.
[31] Zhou J T, Tsang I W, Pan S J, et al. Heterogeneous domain adaptation for multiple classes[C]//Proceedings of the 17th International Conference on Artificial Intelligence and Stati-stics, Reykjavik, Apr 22-25, 2014: 1095-1103.
[32] Tsai Y H, Yeh Y, Wang Y F. Learning cross-domain land-marks for heterogeneous domain adaptation[C]//Proceed-ings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, Jun 27-30, 2016. Washing-ton: IEEE Computer Society, 2016: 5081-5090.
[33] Chen W Y, Hsu T H, Tsai Y H, et al. Transfer neural trees for heterogeneous domain adaptation[C]//LNCS 9909: Pro-ceedings of the 2016 European Conference on Computer Vision, Amsterdam, Oct 11-14, 2016. Berlin, Heidelberg:Springer, 2016: 399-414.
[34] Wang X S, Ma Y T, Cheng Y H, et al. Rodrigues, hetero-geneous domain adaptation network based on autoencoder[J]. Journal of Parallel and Distributed Computing, 2017, 117: 281-291.
[35] Xu Z J, Sun S L. Multi-view transfer learning with Adab-oost[C]//Proceedings of the 23rd IEEE International Confer-ence on Tools with Artificial Intelligence, Boca Raton, Nov 7-9, 2011: 399-402.
[36] Yang P, Gao W. Multi-view discriminant transfer learning [C]//Proceedings of the 23rd International Joint Conference on Artificial Intelligence, Beijing, Aug 3-9, 2013. Menlo Park: AAAI, 2013: 1848-1854.
[37] Hoffman J, Kulis B, Darrell T, et al. Discovering latent dom-ains for multisource domain adaptation[C]//LNCS 7573: Pro-ceedings of the 12th European Conference on Computer Vision, Florence, Oct 7-13, 2012: 702-715.
[38] Long M S, Wang J M, Ding G G, et al. Transfer feature learning with joint distribution adaptation[C]//Proceedings of the 2013 IEEE International Conference on Computer Vision, Sydney, Dec 1-8, 2013. Washington: IEEE Computer Society, 2013: 2200-2207.
[39] Hardoon D R, Szedmak S, Shawetaylor J. Canonical corre-lation analysis: an overview with application to learning met-hods[J]. Neural Computation, 2004, 16: 2639-2664.
[40] He X F, Niyogi P. Locality preserving projections[C]//Pro-ceedings of the 2003 Advances in Neural Information Proce-ssing Systems. Cambridge: MIT Press, 2003: 153-160.
[41] Belkin M, Niyogi P. Laplacian eigenmaps and spectral tech-niques for embedding and clustering[C]//Proceedings of the 2001 Advances in Neural Information Processing Systems, Vancouver, Dec 3-8, 2001. Cambridge: MIT Press, 2001: 585-591.
[42] Sharma A, Paliwal K K. Linear discriminant analysis for the small sample size problem: an overview[J]. International Jour-nal of Machine Learning and Cybernetics, 2015, 6(3): 443-454.