Abstract
Transfer learning is a problem defined over two domains. These two domains share the same feature space and class label space, but have significantly different distributions. One domain has sufficient labels, named as source domain, and the other domain has few labels, named as target domain. The problem is to learn an effective classifier for the target domain. In this paper, we propose a novel transfer learning method for this problem by learning a partially shared classifier for the target domain, and weighting the source domain data points. We learn some shared subspaces for both the data points of the two domains, and a shared classifier in the shared subspaces. We hope that in the shared subspaces, the distributions of two domain can match each other well, and to match the distributions, we weight the source domain data points with different weighting factors. Moreover, we adapt the shared classifier to each domain by learning different adaptation functions. To learn the subspace transformation matrices, the classifier parameters, and the adaptation parameters, we build an objective function with weighted classification errors, parameter regularization, local reconstruction regularization, and distribution matching. This objective function is minimized by an iterative algorithm. Experiments show its effectiveness over benchmark data sets, including travel destination review data set, face expression data set, spam email data set.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Chen A, Eberle M, Lunt E, Liu S, Leake K, Rudenko M, Hawkins A, Schmidt H (2011) Dual-color fluorescence cross-correlation spectroscopy on a planar optofluidic chip. Lab Chip 11(8):1502–1506
Chen B, Lam W, Tsang I, Wong TL (2013) Discovering low-rank shared concept space for adapting text mining models. IEEE Trans Pattern Anal Mach Intell 35(6):1284–1297
Chu WS, Torre FDL, Cohn JF (2013) Selective transfer machine for personalized facial action unit detection. In: IEEE conference on computer vision and pattern recognition (CVPR), 2013, pp. 3515–3522
Fan X, Malone B, Yuan C (2014) Finding optimal bayesian network structures with constraints learned from data. In: Proceedings of the 30th conference on uncertainty in artificial intelligence (UAI-2014)
Fan X, Tang K (2010) Enhanced maximum auc linear classifier. In: 2010 Seventh international conference on fuzzy systems and knowledge discovery (FSKD), vol 4. IEEE, pp 1540–1544
Fan X, Tang K, Weise T (2011) Margin-based over-sampling method for learning from imbalanced datasets. In: Pacific-Asia conference on knowledge discovery and data mining. Springer, pp 309–320
Fan X, Yuan C (2015) An improved lower bound for bayesian network structure learning. In: Proceedings of the 29th AAAI conference on artificial intelligence (AAAI-2015)
Fan X, Yuan C, Malone B (2014) Tightening bounds for bayesian network structure learning. In: Proceedings of the 28th AAAI conference on artificial intelligence (AAAI-2014)
La L, Guo Q, Cao Q, Wang Y (2014) Transfer learning with reasonable boosting strategy. Neural Comput Appl 24(3–4):807–816
Li W, Duan L, Xu D, Tsang I (2014) Learning with augmented features for supervised and semi-supervised heterogeneous domain adaptation. IEEE Trans Pattern Anal Mach Intell 36(6):1134–1148
Lin F, Wang J, Zhang N, Xiahou J, McDonald N (2016) Multi-kernel learning for multivariate performance measures optimization. Neural Comput Appl 1–13. doi:10.1007/s00521-015-2164-9
Liu S, Hawkins AR, Schmidt H (2016) Optofluidic devices with integrated solid-state nanopores. Microchim Acta 183(4):1275–1287
Liu S, Wall TA, Ozcelik D, Parks JW, Hawkins AR, Schmidt H (2015) Electro-optical detection of single \(\lambda\)-dna. Chem Commun 51(11):2084–2087
Liu S, Yuzvinsky TD, Schmidt H (2013) Effect of fabrication-dependent shape and composition of solid-state nanopores on single nanoparticle detection. ACS Nano 7(6):5621–5627
Liu S, Zhao Y, Parks JW, Deamer DW, Hawkins AR, Schmidt H (2014) Correlated electrical and optical analysis of single nanoparticles and biomolecules on a nanopore-gated optofluidic chip. Nano Lett 14(8):4816–4820
Liu X, Wang J, Yin M, Edwards B, Xu P (2015) Supervised learning of sparse context reconstruction coefficients for data representation and classification. Neural Comput Appl 1–9. doi:10.1007/s00521-015-2042-5
Liu Y, Yang J, Zhou Y, Hu J (2013) Structure design of vascular stents. In: Li S, Qian D (eds) Multiscale simulations and mechanics of biological materials. Wiley, Oxford, pp 301–317
Ma Z, Yang Y, Sebe N, Hauptmann AG (2014) Knowledge adaptation with partiallyshared features for event detectionusing few exemplars. IEEE Trans Pattern Anal Mach Intell 36(9):1789–1802
Peng B, Liu Y, Zhou Y, Yang L, Zhang G, Liu Y (2015) Modeling nanoparticle targeting to a vascular surface in shear flow through diffusive particle dynamics. Nanoscale Res Lett 10(1):1–9
Seera M, Lim C (2013) Transfer learning using the online fuzzy min-max neural network. Neural Comput Appl 25:1–12
Seera M, Lim C (2014) Transfer learning using the online fuzzy minmax neural network. Neural Comput Appl 25(2):469–480
Tahmoresnezhad J, Hashemi S (2016) Visual domain adaptation via transfer feature learning. Knowl Inf Syst 1–21. doi:10.1007/s10115-016-0944-x
Wang H, Wang J (2014) An effective image representation method using kernel classification. In: 2014 IEEE 26th International conference on tools with artificial intelligence (ICTAI). IEEE, pp 853–858
Wang J, Wang H, Zhou Y, McDonald N (2015) Multiple kernel multivariate performance learning using cutting plane algorithm. In: 2015 IEEE International conference on systems, man, and cybernetics (SMC). IEEE, pp 1870–1875
Wang J, Zhou Y, Duan K, Wang JJY, Bensmail H (2015) Supervised cross-modal factor analysis for multiple modal data classification. In: 2015 IEEE International conference on systems, man, and cybernetics (SMC). IEEE, pp 1882–1888
Wang S, Zhou Y, Tan J, Xu J, Yang J, Liu Y (2014) Computational modeling of magnetic nanoparticle targeting to stent surface under high gradient field. Comput Mech 53(3):403–412
Xiao M, Guo Y (2015) Feature space independent semi-supervised domain adaptation via kernel matching. IEEE Trans Pattern Anal Mach Intell 37(1):54–66
Xu J, Yang J, Huang N, Uhl C, Zhou Y, Liu Y (2016) Mechanical response of cardiovascular stents under vascular dynamic bending. Biomed Eng Online 15(1):1
Yang S, Hou C, Zhang C, Wu Y (2013) Robust non-negative matrix factorization via joint sparse and graph regularization for transfer learning. Neural Comput Appl 23(2):541–559
Yang S, Lin M, Hou C, Zhang C, Wu Y (2012) A general framework for transfer sparse subspace learning. Neural Comput Appl 21(7):1801–1817
Zhou Y, Hu W, Peng B, Liu Y (2014) Biomarker binding on an antibody-functionalized biosensor surface: the influence of surface properties, electric field, and coating density. J Phys Chem C 118(26):14586–14594
Zhou Y, Sohrabi S, Tan J, Liu Y (2016) Mechanical properties of nanoworm assembled by dna and nanoparticle conjugates. J Nanosci Nanotechnol 16(6):5447–5456
Acknowledgments
This research was supported by National Natural Science Foundation (71173062, 71203047), Key Program for Science and Technology Research of Heilongjiang Province (GB14D201), and University Academic Innovation Team Construction Plan of Philosophy and Social Sciences in Heilongjiang Province (TD201203).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Wang, H., Xu, A., Wang, S. et al. Cross domain adaptation by learning partially shared classifiers and weighting source data points in the shared subspaces. Neural Comput & Applic 29, 237–248 (2018). https://doi.org/10.1007/s00521-016-2541-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-016-2541-z