Abstract
In this paper, we apply Restricted Boltzmann Machine and Subspace Restricted Boltzmann Machine to domain adaptation. Moreover, we train these models using the Perturb-and-MAP approach to draw approximate sample from the Gibbs distribution. We evaluate our approach on domain adaptation task between two image corpora: MNIST and Handwritten Character Recognition dataset.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
\(\mathbb {I}[\cdot ]\) denotes the indicator function, and \(\odot \) is the element-wise multiplication.
- 2.
The original HCR is scaled to fit MNIST images.
- 3.
RBM and sRBM trained with contrastive divergence (CD) and PD are denoted by RBM-CD and sRBM-CD, and RBM-PD and sRBM-PD, respectively.
References
Ben-David, S., Blitzer, J., Crammer, K., Kulesza, A., Pereira, F., Vaughan, J.W.: A theory of learning from different domains. Mach. Learn. 79(1–2), 151–175 (2010)
Bengio, Y.: Learning deep architectures for AI. Found. Trends Mach. Learn. 2(1), 1–127 (2009)
Chelba, C., Acero, A.: Adaptation of maximum entropy capitalizer: little data can help a lot. Comput. Speech Lang. 20(4), 382–399 (2006)
Daumé III., H., Frustratingly easy domain adaptation. arXiv preprint arXiv: 0907.1815 (2009)
Daume III, H., Marcu, D.: Domain adaptation for statistical classifiers. J. Artif. Intell. Res. 26, 101–126 (2006)
Ganin, Y., Lempitsky, V.: Unsupervised domain adaptation by backpropagation. In: Proceedings of The 32nd International Conference on Machine Learning, pp. 1180–1189 (2015)
Glorot, X., Bordes, A., Bengio, Y.: Domain adaptation for large-scale sentiment classification: a deep learning approach. In: Proceedings of the 28th International Conference on Machine Learning (ICML 2011), pp. 513–520 (2011
Hazan, T., Jaakkola, T.: On the partition function and random maximum a-posteriori perturbations. In: ICML (2012)
Hazan, T., Maji, S., Jaakkola, T.: On sampling from the gibbs distribution with random maximum a-posteriori perturbations. In: NIPS, pp. 1268–1276 (2013)
Hinton, G.: Training products of experts by minimizing contrastive divergence. Neural Comput. 14(8), 1771–1800 (2002)
Hinton, G.E., Sejnowski, T.J.: Learning and relearning in Boltzmann machines. Parallel Distrib. Process. Explor. Microstruct. Cogn. 1, 282–317 (1986)
Jiang, J.: A literature survey on domain adaptation of statistical classifiers. Technical Report http://sifaka.cs.uiuc.edu/jiang4/domain_adaptation/survey/da_survey.pdf
LeCun, Y., Bottou, L., Bengio, Y., Haffner, P.: Gradient-based learning applied to document recognition. Proc. IEEE 86(11), 2278–2324 (1998)
Li, X., Bilmes, J.: A bayesian divergence prior for classiffier adaptation. In: International Conference on Artificial Intelligence and Statistics, pp. 275–282 (2007)
Liu, A., Ziebart, B.: Robust classification under sample selection bias. In: Advances in Neural Information Processing Systems, pp. 37–45 (2014)
Long, M., Wang, J., Ding, G., Pan, S.J., Yu, P.S.: Adaptation regularization: a general framework for transfer learning. IEEE Trans. Knowl. Data Eng. 26(5), 1076–1089 (2014)
Orabona, F., Hazan, T., Sarwate, A., Jaakkola, T.: On measure concentration of random maximum a-posteriori perturbations. In: ICML, pp. 432–440 (2014)
Papandreou, G., Yuille, A., Perturb-and-map random fields: using discrete optimization to learn and sample from energy models. In: ICCV, pp. 193–200 (2011)
Ravanbakhsh, S., Greiner, R., Frey, B.: Machine, Training Restricted Boltzmann by Perturbation. arXiv preprint arXiv: 1405.1436 (2014)
Sejnowski, T.: Higher-order Boltzmann machines. In: AIP Conference Proceedings, vol. 151, pp. 398–403 (1986)
Smolensky, P.: Information processing in dynamical systems: foundations of harmony theory. In: Parallel Distributed Processing: Explorations in the Microstructure of Cognition, vol. 1, pp. 194–281. MIT Press (1986)
Tomczak, J.M.: On some properties of the low-dimensional Gumbel perturbations in the Perturb-and-MAP model. Stat. Probab. Lett. 115, 8–15 (2016). http://dx.doi.org/10.1016/j.spl.2016.03.019
Tomczak, J.M., Gonczarek, A.: Learning Invariant Features Using Subspace Restricted Boltzmann Machine. Neural Process. Lett. 1–10 (2016). doi:10.1007/s11063-016-9519-9
Tommasi, T., Orabona, F., Caputo, B.: Learning categories from few examples with multi model knowledge transfer. IEEE Trans. Pattern Anal. Mach. Intell. 36(5), 928–941 (2014)
Van der Maaten, L.: A new benchmark dataset for handwritten character recognition, pp. 2–5. Tilburg University (2009)
Yosinski, J., Clune, J., Bengio, Y., Lipson, H.: How transferable are features in deep neural networks? In: Advances in Neural Information Processing Systems, pp. 3320–3328 (2014)
Acknowledgments
The work presented in the paper is partially co-financed by the Ministry of Science and Higher Education in Poland.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Zaręba, S., Kocot, M., Tomczak, J.M. (2017). Domain Adaptation for Image Analysis: An Unsupervised Approach Using Boltzmann Machines Trained by Perturbation. In: Świątek, J., Tomczak, J. (eds) Advances in Systems Science. ICSS 2016. Advances in Intelligent Systems and Computing, vol 539. Springer, Cham. https://doi.org/10.1007/978-3-319-48944-5_2
Download citation
DOI: https://doi.org/10.1007/978-3-319-48944-5_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-48943-8
Online ISBN: 978-3-319-48944-5
eBook Packages: EngineeringEngineering (R0)