×

Signal classification with a point process distance on the space of persistence diagrams. (English) Zbl 1416.94030

Summary: In this paper, we consider the problem of signal classification. First, the signal is translated into a persistence diagram through the use of delay-embedding and persistent homology. Endowing the data space of persistence diagrams with a metric from point processes, we show that it admits statistical structure in the form of Fréchet means and variances and a classification scheme is established. In contrast with the Wasserstein distance, this metric accounts for changes in small persistence and changes in cardinality. The classification results using this distance are benchmarked on both synthetic data and real acoustic signals and it is demonstrated that this classifier outperforms current signal classification techniques.

MSC:

94A12 Signal theory (characterization, reconstruction, filtering, etc.)
62H30 Classification and discrimination; cluster analysis (statistical aspects)
55N35 Other homology theories in algebraic topology

Software:

Ripser; TDA; GitHub; Dionysus
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Adcock, A.; Carlsson, E.; Carlsson, G., The ring of algebraic functions on persistence bar codes, Homol Homotopy Appl, 18, 381-402, (2016) · Zbl 1420.55017 · doi:10.4310/HHA.2016.v18.n1.a21
[2] Adler, RJ; Bobrowski, O.; Weinberger, S., Crackle: the homology of noise, Discrete Comput Geom, 52, 680-704, (2014) · Zbl 1350.60010 · doi:10.1007/s00454-014-9621-6
[3] Azimi-Sadjadi MR, Yang Y, Srinivasan S (2007) Acoustic classification of battlefield transient events using wavelet subband features. In: Proceedings of SPIE defense and security symposium, p 6562
[4] Bampasidou M, Gentimis T (2014) Modeling collaborations with persistent homology. arXiv preprint arXiv:1403.5346
[5] Bauer U (2015) Ripser. https://github.com/Ripser/ripser
[6] Bogert BP, Healy MJ, Tukey JW (1963) The quefrency alanysis of time series for echoes: cepstrum, pseudo-autocovariance, cross-cepstrum and saphe cracking. In: Proceedings of the symposium on time series analysis, chapter, vol 15, pp 209-243
[7] Bubenik, P., Statistical topological data analysis using persistence landscapes, J Mach Learn Res, 16, 77-102, (2015) · Zbl 1337.68221
[8] Carlsson, G., Topology and data, Bull Am Math Soc, 46, 255-308, (2009) · Zbl 1172.62002 · doi:10.1090/S0273-0979-09-01249-X
[9] Chazal F, Cohen-Steiner D, Glisse M, Guibas LJ, Oudot SY (2009) Proximity of persistence modules and their diagrams. In: Proceedings of the twenty-fifth annual symposium on Computational geometry. ACM, pp 237-246 · Zbl 1380.68387
[10] Cohen-Steiner, D.; Edelsbrunner, H.; Harer, J.; Mileyko, Y., Lipschitz functions have \(L_p\)-stable persistence, Found Comput Math, 10, 127-139, (2010) · Zbl 1192.55007 · doi:10.1007/s10208-010-9060-6
[11] Dhanalakshmi, P.; Palanivel, S.; Ramalingam, V., Classification of audio signals using SVM and RBFNN, Expert Syst Appl, 36, 6069-6075, (2009) · doi:10.1016/j.eswa.2008.06.126
[12] Edelsbrunner H, Harer J (2010) Computational topology: an introduction. American Mathematical Society, Providence · Zbl 1193.55001
[13] Emrani, S.; Gentimis, T.; Krim, H., Persistent homology of delay embeddings and its application to wheeze detection, IEEE Signal Process Lett, 21, 459-463, (2015) · doi:10.1109/LSP.2014.2305700
[14] Fasy BT, Kim J, Lecci F, Maria C, Rouvreau V (2015) The included GUDHI is authored by Clement Maria PbUBMK Dionysus by Dmitriy Morozov, Reininghaus J Tda: statistical tools for topological data analysis r package version 1.4.1. https://CRAN.R-project.org/package=TDA
[15] Garrett, D.; Peterson, DA; Anderson, CW; Thaut, MH, Comparison of linear, nonlinear, and feature selection methods for eeg signal classification, IEEE Trans Neural Syst Rehabil Eng, 11, 141-166, (2003) · doi:10.1109/TNSRE.2003.814441
[16] Hatcher A (2002) Algebraic topology. Cambridge University Press, Cambridge · Zbl 1044.55001
[17] Kerber M, Morozov D, Nigmetov A (2016) Geometry helps to compare persistence diagrams. In: Proceedings of the eighteenth workshop on algorithm engineering and experiments, pp 103-112 · Zbl 1414.68129
[18] Krim, H.; Gentimis, T.; Chintakunta, H., Discovering the whole by the coarse: a topological paradigm for data analysis, IEEE Signal Process Mag, 33, 95-104, (2016) · doi:10.1109/MSP.2015.2510703
[19] Kuhn, HW, The Hungarian method for the assignment problem, Naval Res Log Q, 2, 83-87, (1955) · doi:10.1002/nav.3800020109
[20] Law K, Stewart A, Zygalakis K (2015) Data assimilation: a mathematical introduction. Springer, Berlin · Zbl 1353.60002 · doi:10.1007/978-3-319-20325-6
[21] Lum, PY; Singh, G.; Lehman, A.; Ishkanov, T.; Vejdemo-Johansson, M.; Alagappan, M.; Carlsson, J.; Carlsson, G., Extracting insights from the shape of complex data using topology, Sci Rep, 3, 1236, (2013) · doi:10.1038/srep01236
[22] Maroulas, V.; Nebenführ, A., Tracking rapid intracellular movements: a Bayesian random set approach, Ann Appl Stat, 9, 926-949, (2015) · Zbl 1454.62363 · doi:10.1214/15-AOAS819
[23] Mileyko, Y.; Mukherjee, S.; Harer, J., Probability measures on the space of persistence diagrams, Inverse Problems, 27, 124007, (2011) · Zbl 1247.68310 · doi:10.1088/0266-5611/27/12/124007
[24] Nicolau, M.; Levine, A.; Carlsson, G., Topology based data analysis identifies a subgroup of breast cancers with a unique mutational profile and excellent survival, Proc Nat Acad Sci, 108, 7265-7270, (2011) · doi:10.1073/pnas.1102826108
[25] Oppenheim, AV; Schafer, RW, From frequency to quefrency: a history of the cepstrum, IEEE Signal Process Mag, 21, 95-106, (2004) · doi:10.1109/MSP.2004.1328092
[26] Reininghaus J, Huber S, Bauer U, Kwitt R (2015) A stable multi-scale kernel for topological machine learning. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 4741-4748
[27] Robins, V.; Turner, K., Principal component analysis of persistent homology rank functions with case studies of spatial point patterns, sphere packing and colloids, Physica D, 334, 99-117, (2016) · doi:10.1016/j.physd.2016.03.007
[28] Schuhmacher, D.; Vo, B.; Vo, B., A consistent metric for performance evaluation of multi-object filters, IEEE Trans Signal Process, 56, 3447-3457, (2008) · Zbl 1390.94399 · doi:10.1109/TSP.2008.920469
[29] Seversky LM, Davis S, Berger M (2016) On time-series topological data analysis: new data and opportunities. In: The IEEE conference on computer vision and pattern recognition, pp 59-67
[30] Sherwin, J.; Sajda, P., Musical experts recruit action-related neural structures in harmonic anomaly detection: evidence for embodied cognition in expertise, Brain Cogn, 83, 190-202, (2013) · doi:10.1016/j.bandc.2013.07.002
[31] Srinivas U, Nasrabadi NM, Monga V (2013) Graph-based multi-sensor fusion for acoustic signal classification. In: 2013 IEEE international conference on acoustics, speech and signal processing (ICASSP), pp 261-265
[32] Takens, Floris, Detecting strange attractors in turbulence, 366-381, (1981), Berlin, Heidelberg · Zbl 0513.58032
[33] Turner, K.; Mileyko, Y.; Mukherjee, S.; Harer, J., Fréchet means for distributions of persistence diagrams, Discrete Comput Geom, 52, 44-70, (2014) · Zbl 1296.68182 · doi:10.1007/s00454-014-9604-7
[34] Venkataraman V, Ramamurthy KN, Turaga P (2016) Persistent homology of attractors for action recognition. In: 2016 IEEE international conference on image processing (ICIP), pp 4150-4154
[35] Xia, K.; Wei, GW, Persistent homology analysis of protein structure, flexibility, and folding, Int J Numer Methods Biomed Eng, 30, 814-844, (2014) · doi:10.1002/cnm.2655
[36] Zhang H, Nasrabadi NM, Huang TS, Zhang Y (2011) Transient acoustic signal classification using joint sparse representation. In: 2011 IEEE international conference on acoustics, speech and signal processing (ICASSP), pp 2220-2223
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.