ABSTRACT
We introduce Deep Reasoning Networks (DRNets), an end-to-end framework that combines deep learning with constraint reasoning for solving pattern de-mixing problems, typically in an unsupervised or very-weakly-supervised setting. DRNets exploit problem structure and prior knowledge by tightly combining constraint reasoning with stochastic-gradient-based neural network optimization. Our motivating task is from materials discovery and concerns inferring crystal structures of materials from X-ray diffraction data (Crystal-Structure-Phase-Mapping). Given the complexity of its underlying scientific domain, we start by introducing DRNets on an analogous but much simpler task: de-mixing overlapping handwritten Sudokus (Multi-MNIST-Sudoku). On Multi-MNIST-Sudoku, DRNets almost perfectly recovered the mixed Sudokus' digits, with 100% digit accuracy, outperforming the supervised state-of-the-art MNIST de-mixing models. On Crystal-Structure-Phase-Mapping, DRNets significantly outperform the state of the art and experts' capabilities, recovering more precise and physically meaningful crystal structures.
- Bai, J., Bjorck, J., Xue, Y., Suram, S. K., Gregoire, J., and Gomes, C. Relaxation methods for constrained matrix factorization problems: solving the phase mapping problem in materials discovery. In International Conference on AI and OR Techniques in Constraint Programming for Combinatorial Optimization Problems, pp. 104-112. Springer, 2017.Google Scholar
Cross Ref
- Bai, J., Ament, S., Perez, G., Gregoire, J., and Gomes, C. An efficient relaxed projection method for constrained non-negative matrix factorization with application to the phase-mapping problem in materials science. In International Conference on the Integration of Constraint Programming, Artificial Intelligence, and Operations Research, pp. 52-62. Springer, 2018.Google ScholarCross Ref
- Ermon, S., Le Bras, R., Suram, S. K., Gregoire, J. M., Gomes, C. P., Selman, B., and Van Dover, R. B. Pattern decomposition with complex combinatorial constraints: Application to materials discovery. In Twenty-Ninth AAAI Conference on Artificial Intelligence, 2015.Google Scholar
- Ganchev, K., Gillenwater, J., Taskar, B., et al. Posterior regularization for structured latent variable models. Journal of Machine Learning Research, 11(Jul):2001-2049, 2010.Google Scholar
- Garcez, A. d., Gori, M., Lamb, L. C., Serafini, L., Spranger, M., and Tran, S. N. Neural-symbolic computing: An effective methodology for principled integration of machine learning and reasoning. arXiv preprint arXiv:1905.06088, 2019.Google Scholar
- Gomes, C. P., Selman, B., Kautz, H., et al. Boosting combinatorial search through randomization. AAAI/IAAI, 98: 431-437, 1998.Google Scholar
- Gomes, C. P., Bai, J., Xue, Y., Björck, J., Rappazzo, B., Ament, S., Bernstein, R., Kong, S., Suram, S. K., van Dover, R. B., et al. Crystal: a multi-agent ai system for automated mapping of materials' crystal structures. MRS Communications, pp. 1-9, 2019.Google Scholar
- He, K., Zhang, X., Ren, S., and Sun, J. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770-778, 2016.Google ScholarCross Ref
- Hinton, G. E., Ghahramani, Z., and Teh, Y. W. Learning to parse images. In Advances in neural information processing systems, pp. 463-469, 2000.Google Scholar
- Hu, Z., Ma, X., Liu, Z., Hovy, E., and Xing, E. Harnessing deep neural networks with logic rules. arXiv preprint arXiv:1603.06318, 2016a.Google Scholar
- Hu, Z., Yang, Z., Salakhutdinov, R., and Xing, E. Deep neural networks with massive learned knowledge. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pp. 1670-1679, 2016b.Google ScholarCross Ref
- Kingma, D. and Ba, J. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014.Google Scholar
- Le Bras, R., Bernstein, R., Gregoire, J. M., Suram, S. K., Gomes, C. P., Selman, B., and Van Dover, R. B. Challenges in materials discovery-synthetic generator and real datasets. In Twenty-Eighth AAAI Conference on Artificial Intelligence, 2014.Google Scholar
- LeCun, Y., Bottou, L., Bengio, Y., Haffner, P., et al. Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11):2278-2324, 1998.Google ScholarDigital Library
- Lindsay, B. G. Mixture models: theory, geometry and applications. In NSF-CBMS regional conference series in probability and statistics, pp. i-163. JSTOR, 1995.Google Scholar
- Long, C., Bunker, D., Li, X., Karen, V., and Takeuchi, I. Rapid identification of structural phases in combinatorial thin-film libraries using x-ray diffraction and non-negative matrix factorization. Review of Scientific Instruments, 80(10):103902, 2009.Google ScholarCross Ref
- Manhaeve, R., Dumancic, S., Kimmig, A., Demeester, T., and De Raedt, L. Deepproblog: Neural probabilistic logic programming. In Advances in Neural Information Processing Systems, pp. 3749-3759, 2018.Google Scholar
- Mirza, M. and Osindero, S. Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784, 2014.Google Scholar
- Nair, V. and Hinton, G. E. Rectified linear units improve restricted boltzmann machines. In Proceedings of the 27th international conference on machine learning (ICML-10), pp. 807-814, 2010.Google ScholarDigital Library
- Pathak, D., Krahenbuhl, P., and Darrell, T. Constrained convolutional neural networks for weakly supervised segmentation. In Proceedings of the IEEE international conference on computer vision, pp. 1796-1804, 2015.Google ScholarDigital Library
- Robbins, H. and Monro, S. A stochastic approximation method. In Herbert Robbins Selected Papers, pp. 102- 109. Springer, 1985.Google ScholarCross Ref
- Sabour, S., Frosst, N., and Hinton, G. E. Dynamic routing between capsules. In Advances in neural information processing systems, pp. 3856-3866, 2017.Google ScholarDigital Library
- Shivhare, R. and Kumar, C. A. On the cognitive process of abstraction. Procedia Computer Science, 89:243-252, 2016.Google Scholar
- Stanev, V., Vesselinov, V. V., Kusne, A. G., Antoszewski, G., Takeuchi, I., and Alexandrov, B. S. Unsupervised phase mapping of x-ray diffraction data by nonnegative matrix factorization integrated with custom clustering. npj Computational Materials, 4(1):43, 2018.Google Scholar
- Wang, P.-W., Donti, P. L., Wilder, B., and Kolter, Z. Satnet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver. arXiv preprint arXiv:1905.12149, 2019.Google Scholar
- Xu, J., Zhang, Z., Friedman, T., Liang, Y., and Broeck, G. V. d. A semantic loss function for deep learning with symbolic knowledge. arXiv preprint arXiv:1711.11157, 2017.Google Scholar
- Xue, Y., Bai, J., Le Bras, R., Rappazzo, B., Bernstein, R., Bjorck, J., Longpre, L., Suram, S. K., van Dover, R. B., Gregoire, J., et al. Phase-mapper: an ai platform to accelerate high throughput materials discovery. In Twenty-Ninth IAAI Conference, 2017.Google Scholar
- You, J., Ying, R., Ren, X., Hamilton, W. L., and Leskovec, J. Graphrnn: Generating realistic graphs with deep autoregressive models. arXiv preprint arXiv:1802.08773, 2018.Google Scholar
- Zhang, N., Yan, J., and Zhou, Y. Weakly supervised audio source separation via spectrum energy preserved wasserstein learning. arXiv preprint arXiv:1711.04121, 2017.Google Scholar
- Zhou, X., Sun, X., Zhang, W., Liang, S., and Wei, Y. Deep kinematic pose regression. In European Conference on Computer Vision, pp. 186-201. Springer, 2016.Google ScholarCross Ref
- Zhou, X., Huang, Q., Sun, X., Xue, X., and Wei, Y. Weakly-supervised transfer for 3d human pose estimation in the wild. In IEEE International Conference on Computer Vision, volume 206, pp. 3, 2017.Google Scholar
Index Terms
(auto-classified)Deep reasoning networks for unsupervised pattern de-mixing with constraint reasoning
Comments