Towards More Reliable Transfer Learning

  • Zirui WangEmail author
  • Jaime Carbonell
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11052)


Multi-source transfer learning has been proven effective when within-target labeled data is scarce. Previous work focuses primarily on exploiting domain similarities and assumes that source domains are richly or at least comparably labeled. While this strong assumption is never true in practice, this paper relaxes it and addresses challenges related to sources with diverse labeling volume and diverse reliability. The first challenge is combining domain similarity and source reliability by proposing a new transfer learning method that utilizes both source-target similarities and inter-source relationships. The second challenge involves pool-based active learning where the oracle is only available in source domains, resulting in an integrated active transfer learning framework that incorporates distribution matching and uncertainty sampling. Extensive experiments on synthetic and two real-world datasets clearly demonstrate the superiority of our proposed methods over several baselines including state-of-the-art transfer learning methods. Code related to this paper is available at:

Supplementary material

478890_1_En_47_MOESM1_ESM.pdf (180 kb)
Supplementary material 1 (pdf 180 KB)


  1. 1.
    Balcan, M., Hanneke, S., Vaughan, J.W.: The true sample complexity of active learning. Mach. Learn. 80(2), 111–139 (2010)MathSciNetCrossRefGoogle Scholar
  2. 2.
    Ben-David, S., Blitzer, J., Crammer, K., Pereira, F.: Analysis of representations for domain adaptation. In: NIPS, pp. 137–144 (2007)Google Scholar
  3. 3.
    Blitzer, J., Dredze, M., Pereira, F., et al.: Biographies, bollywood, boom-boxes and blenders: domain adaptation for sentiment classification. In: ACL, pp. 440–447 (2007)Google Scholar
  4. 4.
    Blitzer, J., Crammer, K., Kulesza, A., Pereira, F., Wortman, J.: Learning bounds for domain adaptation. In: NIPS, pp. 129–136 (2008)Google Scholar
  5. 5.
    Cavallanti, G., Cesa-Bianchi, N., Gentile, C.: Linear algorithms for online multitask classification. J. Mach. Learn. Res. 11(Oct), 2901–2934 (2010)MathSciNetzbMATHGoogle Scholar
  6. 6.
    Chattopadhyay, R., Fan, W., Davidson, I., Panchanathan, S., Ye, J.P.: Joint transfer and batch-mode active learning. In: ICML, pp. 253–261 (2013)Google Scholar
  7. 7.
    Duan, L.X., Tsang, I.W., Xu, D., Chua, T.: Domain adaptation from multiple sources via auxiliary classifiers. In: ICML, pp. 289–296 (2009)Google Scholar
  8. 8.
    Gretton, A., Borgwardt, K.M., Rasch, M., Scholkopf, B., Smola, A.J.: A kernel method for the two-sample-problem. In: NIPS, pp. 513–520 (2007)Google Scholar
  9. 9.
    Gretton, A., et al.: Optimal kernel choice for large-scale two-sample tests. In: NIPS, pp. 1205–1213 (2012)Google Scholar
  10. 10.
    Halko, N., Martinsson, P.G., Tropp, J.A.: Finding structure with randomness: probabilistic algorithms for constructing approximate matrix decompositions. SIAM Rev. 53(2), 217–288 (2011)MathSciNetCrossRefGoogle Scholar
  11. 11.
    Huang, J.Y., Gretton, A., Borgwardt, K.M., Scholkopf, B., Smola, A.J.: Correcting sample selection bias by unlabeled data. In: NIPS, pp. 601–608 (2007)Google Scholar
  12. 12.
    Huang, S.J., Chen, S.: Transfer learning with active queries from source domain. In: IJCAI, pp. 1592–1598 (2016)Google Scholar
  13. 13.
    Kale, D., Ghazvininejad, M., Ramakrishna, A., He, J.R., Liu, Y.: Hierarchical active transfer learning. In: SIAM International Conference on Data Mining, pp. 514–522 (2015)Google Scholar
  14. 14.
    Konyushkova, K., Sznitman, R., Fua, P.: Learning active learning from data. In: NIPS, pp. 4228–4238 (2017)Google Scholar
  15. 15.
    Long, M.S., Wang, J.M., Jordan, M.I.: Deep transfer learning with joint adaptation networks. In: ICML, pp. 2208–2217 (2017)Google Scholar
  16. 16.
    Luo, P., Zhuang, F.Z., Xiong, H., Xiong, Y.H., He, Q.: Transfer learning from multiple source domains via consensus regularization. In: ACM Conference on Information and Knowledge Management, pp. 103–112 (2008)Google Scholar
  17. 17.
    Moon, S., Carbonell, J.: Completely heterogeneous transfer learning with attention-what and what not to transfer. In: IJCAI, pp. 2508–2514 (2017)Google Scholar
  18. 18.
    Murugesan, K., Carbonell, J.: Active learning from peers. In: NIPS, pp. 7011–7020 (2017)Google Scholar
  19. 19.
    Murugesan, K., Liu, H.X., Carbonell, J., Yang, Y.M.: Adaptive smoothed online multi-task learning. In: NIPS, pp. 4296–4304 (2016)Google Scholar
  20. 20.
    Nguyen, H.T., Smeulders, A.: Active learning using pre-clustering. In: ICML, p. 79 (2004)Google Scholar
  21. 21.
    Pan, S.J., Ni, X., Sun, J., Yang, Q., Chen, Z.: Cross-domain sentiment classification via spectral feature alignment. ACM Trans. Knowl. Discov. Data (TKDD) 8(3), 12 (2014)Google Scholar
  22. 22.
    Pan, S.J., Tsang, I.W., Kwok, J.T., Yang, Q.: Domain adaptation via transfer component analysis. IEEE Trans. Neural Networks 22(2), 199–210 (2011)CrossRefGoogle Scholar
  23. 23.
    Pan, S.J., Yang, Q.: A survey on transfer learning. IEEE Trans. Knowl. Data Eng. 22(10), 1345–1359 (2010)CrossRefGoogle Scholar
  24. 24.
    Pardoe, D., Stone, P.: Boosting for regression transfer. In: ICML, pp. 863–870 (2010)Google Scholar
  25. 25.
    Raina, R., Battle, A., Lee, H., Packer, B., Ng, A.Y.: Self-taught learning: transfer learning from unlabeled data. In: ICML, pp. 759–766 (2007)Google Scholar
  26. 26.
    Settles, B.: Active learning literature survey. Technical report, University of Wisconsin, Madison (2010)Google Scholar
  27. 27.
    Settles, B., Craven, M.: An analysis of active learning strategies for sequence labeling tasks. In: EMNLP, pp. 1070–1079 (2008)Google Scholar
  28. 28.
    Sun, Q., Chattopadhyay, R., Panchanathan, S., Ye, J.P.: A two-stage weighting framework for multi-source domain adaptation. In: NIPS, pp. 505–513 (2011)Google Scholar
  29. 29.
    Tong, S., Koller, D.: Support vector machine active learning with applications to text classification. J. Mach. Learn. Res. 2(Nov), 45–66 (2001)zbMATHGoogle Scholar
  30. 30.
    Wang, X.Z., Huang, T.K., Schneider, J.: Active transfer learning under model shift. In: ICML, pp. 1305–1313 (2014)Google Scholar
  31. 31.
    Wei, P.F., Sagarna, R., Ke, Y.P., Ong, Y.S., Goh, C.K.: Source-target similarity modelings for multi-source transfer gaussian process regression. In: ICML, pp. 3722–3731 (2017)Google Scholar
  32. 32.
    Wei, Y., Zheng, Y., Yang, Q.: Transfer knowledge between cities. In: KDD, pp. 1905–1914 (2016)Google Scholar
  33. 33.
    Yang, J., Yan, R., Hauptmann, A.G.: Cross domain video concept detection using adaptive SVMs. In: ACM International Conference on Multimedia, pp. 188–197 (2007)Google Scholar
  34. 34.
    Yang, L., Hanneke, S., Carbonell, J.: A theory of transfer learning with applications to active learning. Mach. Learn. 90(2), 161–189 (2013)MathSciNetCrossRefGoogle Scholar
  35. 35.
    Zhang, L., Zuo, W.M., Zhang, D.: LSDT: latent sparse domain transfer learning for visual adaptation. IEEE Trans. Image Process. 25(3), 1177–1191 (2016)MathSciNetCrossRefGoogle Scholar
  36. 36.
    Zhang, Y., Yeung, D.Y.: A regularization approach to learning task relationships in multitask learning. In: WWW, pp. 751–760 (2010)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Language Technologies InstituteCarnegie Mellon UniversityPittsburghUSA

Personalised recommendations