Sliced Wasserstein kernel for persistence diagrams

ABSTRACT

Persistence diagrams (PDs) play a key role in topological data analysis (TDA), in which they are routinely used to describe topological properties of complicated shapes. PDs enjoy strong stability properties and have proven their utility in various learning contexts. They do not, however, live in a space naturally endowed with a Hilbert structure and are usually compared with non-Hilbertian distances, such as the bottleneck distance. To incorporate PDs in a convex learning pipeline, several kernels have been proposed with a strong emphasis on the stability of the resulting RKHS distance w.r.t. perturbations of the PDs. In this article, we use the Sliced Wasserstein approximation of the Wasserstein distance to define a new kernel for PDs, which is not only provably stable but also discriminative (with a bound depending on the number of points in the PDs) w.r.t. the first diagram distance between PDs. We also demonstrate its practicality, by developing an approximation technique to reduce kernel computation time, and show that our proposal compares favorably to existing kernels for PDs on several benchmarks.

References

1. Adams, H., Emerson, T., Kirby, M., Neville, R., Peterson, C., Shipman, P., Chepushtanova, S., Hanson, E., Motta, F., and Ziegelmeier, L. Persistence Images: A Stable Vector Representation of Persistent Homology. Journal Machine Learning Research, 18(8):1-35, 2017.
2. Bauer, U. and Lesnick, M. Induced matchings and the algebraic stability of persistence barcodes. Journal of Computational Geometry, 6(2):162-191, 2015.
3. Berg, C., Christensen, J., and Ressel, P. Harmonic Analysis on Semigroups: Theory of Positive Definite and Related Functions. Springer, 1984.
4. Bubenik, P. Statistical Topological Data Analysis using Persistence Landscapes. Journal Machine Learning Research, 16:77-102, 2015.
5. Carlsson, G. Topology and data. Bulletin American Mathematical Society, 46:255-308, 2009.
6. Carrière, M., Oudot, S., and Ovsjanikov, M. Stable Topological Signatures for Points on 3D Shapes. In Proceedings 13th Symposium Geometry Processing, 2015.
7. Carrière, M., Cuturi, M., and Oudot, S. Sliced Wasserstein Kernel for Persistence Diagrams. CoRR, abs/1706.03358, 2017.
8. Chang, C. and Lin, C. LIBSVM: A library for support vector machines. ACM Transactions on Intelligent Systems and Technology, 2:27:1-27:27, 2011. Software available at http://www.csie.ntu.edu.tw/~cjlin/libsvm.
9. Chazal, F., Cohen-Steiner, D., Glisse, M., Guibas, L., and Oudot, S. Proximity of persistence modules and their diagrams. In Proceedings 25th Symposium Computational Geometry, pp. 237-246, 2009a.
10. Chazal, F., Cohen-Steiner, D., Guibas, L., Memoli, F., and Oudot, S. Gromov-Hausdorff Stable Signatures for Shapes using Persistence. Computer Graphics Forum, pp. 1393-1403, 2009b.
11. Chazal, F., de Silva, V., and Oudot, S. Persistence stability for geometric complexes. Geometriae Dedicata, pp. 1-22, 2013.
12. Chazal, F., de Silva, V., Glisse, M., and Oudot, S. The structure and stability of persistence modules. Springer, 2016.
13. Chen, X., Golovinskiy, A., and Funkhouser, T. A Benchmark for 3D Mesh Segmentation. ACM Trans. Graph., 28(3):73:1-73:12, 2009.
14. Cohen-Steiner, D., Edelsbrunner, H., and Harer, J. Stability of persistence diagrams. Discrete Computational Geometry, 37(1):103-120, 2007.
15. Edelsbrunner, H. and Harer, J. Computational Topology: an introduction. AMS Bookstore, 2010.
16. Edelsbrunner, Herbert and Harer, John. Persistent homology-a survey. Contemporary mathematics, 453:257-282, 2008.
17. Fabio, B. Di and Ferri, M. Comparing persistence diagrams through complex vectors. CoRR, abs/1505.01335, 2015.
18. Guo, Z., Zhang, L., and Zhang, D. A completed modeling of local binary pattern operator for texture classification. IEEE Trans. Image Processing, pp. 1657-1663, 2010.
19. Hertzsch, J.-M., Sturman, R., and Wiggins, S. DNA microarrays: design principles for maximizing ergodic, chaotic mixing. In Small, volume 3, pp. 202-218, 2007.
20. Hiraoka, Y., Nakamura, T., Hirata, A., Escolar, E., Matsue, K., and Nishiura, Y. Hierarchical structures of amorphous solids characterized by persistent homology. In Proceedings National Academy of Science, volume 26, 2016.
21. Kusano, G., Fukumizu, K., and Hiraoka, Y. Persistence Weighted Gaussian Kernel for Topological Data Analysis. In Proceedings 33rd International Conference on Machine Learning, pp. 2004-2013, 2016.
22. Kusano, G., Fukumizu, K., and Hiraoka, Y. Kernel method for persistence diagrams via kernel embedding and weight factor. CoRR, abs/1706.03472, 2017.
23. Kwitt, Roland, Huber, Stefan, Niethammer, Marc, Lin, Weili, and Bauer, Ulrich. Statistical Topological Data Analysis - A Kernel Perspective. In Advances in Neural Information Processing Systems 28, pp. 3070-3078, 2015.
24. Li, C., Ovsjanikov, M., and Chazal, F. Persistence-Based Structural Recognition. In Proceedings Conference Computer Vision Pattern Recognition, pp. 2003-2010, 2014.
25. Morariu, V., Srinivasan, B., Raykar, V., Duraiswami, R., and Davis, L. Automatic online tuning for fast Gaussian summation. In Advances Neural Information Processing Systems 21, pp. 1113-1120, 2009.
26. Ojala, T., Mäenpää, T., Pietikäinen, M., Viertola, J., Kyllönen, J., and Huovinen, S. Outex - new framework for empirical evaluation of texture analysis algorithms. In Proceedings 16th International Conference Pattern Recognition, pp. 701-706, 2002.
27. Oudot, S. Persistence Theory: From Quiver Representations to Data Analysis. American Mathematical Society, 2015.
28. Rabin, J., Peyré, G., Delon, J., and Bernot, M. Wasserstein barycenter and its application to texture mixing. In International Conference Scale Space Variational Methods Computer Vision, pp. 435-446, 2011.
29. Rahimi, A. and Recht, B. Random Features for Large-Scale Kernel Machines. In Advances Neural Information Processing Systems 20, pp. 1177-1184, 2008.
30. Reininghaus, J., Huber, S., Bauer, U., and Kwitt, R. A Stable Multi-Scale Kernel for Topological Machine Learning. CoRR, abs/1412.6821, 2014.
31. Reininghaus, J., Huber, S., Bauer, U., and Kwitt, R. A Stable Multi-Scale Kernel for Topological Machine Learning. In Proceedings Conference Computer Vision Pattern Recognition, 2015.
32. Robins, V. and Turner, K. Principal Component Analysis of Persistent Homology Rank Functions with case studies of Spatial Point Patterns, Sphere Packing and Colloids. Physica D: Nonlinear Phenomena, 334:1-186, 2016.
33. Santambrogio, Filippo. Optimal transport for applied mathematicians. Birkauser, 2015.
34. Singh, G., Memoli, F., Ishkhanov, T., Sapiro, G., Carlsson, G., and Ringach, D. Topological analysis of population activity in visual cortex. Journal of Vision, 8, 2008.
35. The GUDHI Project. GUDHI User and Reference Manual. GUDHI Editorial Board, 2015. URL http://gudhi.gforge.inria.fr/doc/latest/.
36. Villani, C. Optimal transport : old and new. Springer, 2009.
37. Zomorodian, Afra and Carlsson, Gunnar. Computing persistent homology. Discrete & Computational Geometry, 33(2):249-274, 2005.

Index Terms

• Computing methodologies

  • Machine learning