Wykrywanie anomalii w numerycznych cyklach stick-slip przy pomocy algorytmu nienadzorowanego uczenia maszynowego Isolation Forest

• Autorzy: Klejment Piotr

Identyfikatory

DOI

10.32045/PG-2024-050

Warianty tytułu

EN

Detecting anomalies in numerical stick-slip cycles with the unsupervised algorithm isolation forest

• Języki publikacji: PL

Abstrakty

PL

Nienadzorowane uczenie maszynowe kryje w sobie duży, ale wciąż nie do końca wykorzystany potencjał. W niniejszej pracy zastosowano algorytm Isolation Forest pochodzący z rodziny algorytmów nienadzorowanego uczenia maszynowego do wykrywania anomalii w numerycznym modelu cykli stick-slip. Do modelowania numerycznego zastosowano Metodę Elementów Dyskretnych, a w modelu uwzględniono obecność warstwy granularnej. W symulacji zastosowano losowy mechanizm doboru czasu między kolejnymi wydarzenia typu slip, jak i gromadzącego się stress, w efekcie osiągając nieregularny, bardziej realistyczny schemat cykli stick-slip. W każdym z wydarzeń typu slip dokonano rejestracji parametrów opisujących stan układu w danym momencie. Pokazano, że w sposób automatyczny Isolation Forest był w stanie wyizolować anomalie dla każdego ze slip. Miejscem gromadzenia się anomalii były warstwy skrajne i warstwa granularna. Potwierdza to istotną rolę warstwy granularnej w charakterystyce cykli stick-slip. Do oceny wpływu poszczególnych parametrów na działanie algorytmu zastosowano SHapley Additive exPlanations (SHAP).

EN

Unsupervised machine learning has a large but still not fully exploited potential. In this work, the Isolation Forest algorithm, which comes from the family of unsupervised machine learning algorithms, was used to detect anomalies in the numerical model of stick-slip cycles. The Discrete Element Method was used for numerical modeling. The model took into account the presence of a granular layer, which is characteristic of mature faults. A random mechanism for selecting the time between subsequent slip events and the critical strength was used in the simulation, resulting in an irregular, more realistic pattern of stick-slip cycles. In each of the slip events, parameters describing the state of the system at a given moment were recorded. It was shown that Isolation Forest was able to automatically isolate anomalies for each slip. The places where the anomalies accumulated were the outer layers and the granular layer. This confirms the important role of the granular layer in the characteristics of stick-slip cycles. SHapley Additive exPlanations (SHAP) was used to assess the impact of individual parameters on the performance of the algorithm.

Słowa kluczowe

PL

cykl stick-slip; nienadzorowane uczenie maszynowe; Metoda Elementów Dyskretnych

EN

stick-slip cycle; unsupervised machine learning; Discrete Element Method

• Wydawca: Polskie Towarzystwa Geofizyczne ; Komitet Geofizyki PAN
• Czasopismo: Przegląd Geofizyczny
• Rocznik: 2024
• Tom: Z. 3-4
• Strony: 115--133
• Opis fizyczny: Bibliogr. 40 poz., rys., tab., wykr.

Twórcy

autor

Klejment Piotr

• Instytut Geofizyki Polskiej Akademii Nauk

Bibliografia

• 1. Abe S., Boros V., Hancock W., Weatherley D., 2014, ESyS-particle tutorial and user's guide. Version 2.3.1, available online at https://launchpad.net/esys-particle (data access 22.11.2024).
• 2. Airlangga G., 2023, Unsupervised machine learning for seismic anomaly detection: Isolation Forest algorithm application to Indonesian earthquake data, Lebesgue: Jurnal Ilmiah Pendidikan Matematika, Matematika dan Statistika, 4 (3), 1827-1836, DOI: 10.46306/lb.v4i3.479.
• 3. Bai Q., Konietzky H., 2022, Numerical insight into energy partitioning during stick-slip events based on the framework of rate-and-state friction law, Geophysical Research Letters, 49 (9), e2021GL096677, DOI: 10.1029/2021GL096677.
• 4. Bayane I., Leander J., Karoumi R., 2024, An unsupervised machine learning approach for real-time damage detection in bridges, Engineering Structures, 308, DOI: 10.1016/j.engstruct.2024.117971.
• 5. Beroza G.C., Segou M., Mostafa Mousavi S., 2021, Machine learning and earthquake forecasting - next steps, Nature Communications, 12, DOI: 10.1038/s41467-021-24952-6.
• 6. Bolton D.C., Shreedharan S., Riviere J., Marone C., 2020, Acoustic energy release during the laboratory seismic cycle: Insights on laboratory earthquake precursors and prediction, Journal of Geophysical Research: Solid Earth, 125, e2019JB018975, DOI: 10.1029/2019JB018975.
• 7. Celecia A., Figueiredo K., Rodriguez C., Vellasco M., Maldonado E., Silva M.A., Rodrigues A., Nascimento R., Ourofino C., 2021, Unsupervised machine learning applied to seismic interpretation: towards an unsupervised automated interpretation tool, Sensors, 21 (19), DOI: 10.3390/ s21196347.
• 8. Chukwu C., Betts P., Moore D., Munukutla R., Armit R., McLean M., Grose L., 2024, Unsupervised machine learning and depth clusters of Euler deconvolution of magnetic data: a new approach to imaging geological structures, Exploration Geophysics, 55 (3), 223-245, DOI: 10.1080/08123985.2023.2299475.
• 9. Cundall P.A., Strack O.D.L., 1979,, A discrete numerical model for granular assemblies, Geotechnique 29 (1), 47-65, DOI: 10.1680/geot.1979.29.1.47.
• 10. Dahmen K., Ben-Zion Y., Uhl J.A., 2011, A simple analytic theory for the statistics of avalanches in sheared granular materials, Nature Physics, 7, 554-557, DOI: 10.1038/nphys1957.
• 11. Dorostkar O., 2018, Stick-slip dynamics in dry and fluid saturated granular fault gouge investigated by numerical simulations, Doctoral Thesis, ETH Zurich, DOI: 10.3929/ethz-b-000283977.
• 12. Dorostkar O., Carmeliet J., 2018, Potential energy as metric for understanding stick-slip dynamics in sheared granular fault Gguge: a coupled CFD-DEM study, Rock Mechanics and Rock Engineering, 51, 3281-3294, DOI: 10.1007/s00603-018-1457-6.
• 13. Dorostkar O., Carmeliet J., 2019, Grain friction controls characteristics of seismic cycle in faults with granular gouge, Journal of Geophysical Research: Solid Earth, 124 (7), 6475-6489, DOIL 10.1029/2019JB017374.
• 14. Ferdowsi B., 2014, Discrete element modeling of triggered slip in faults with granular gouge: application to dynamic earthquake triggering, Doctoral Dissertation, ETH Zurich, DOI: 10.3929/ ethz-a-010232908.
• 15. Géron A., 2019, Hands-Oon Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition, O'Reilly Media, Inc.
• 16. Gino V.L.S., Negri R.G., Souza F.N., Silva E.A., Bressane A., Mendes T.S.G., Casaca W., 2023, Integrating unsupervised machine intelligence and anomaly detection for spatio-temporal dynamic mapping using remote sensing image series, Sustainability, 15 (6), DOI: 10.3390/su15064725.
• 17. Huang W., 2019, Seismic signal recognition by unsupervised machine learning, Geophysical Journal International, 219 (2), 1163-1180, DOI: 10.1093/gji/ggz366.
• 18. Huang W., Gao K., Feng Y., 2024, Predicting stick-slips in sheared granular fault using machine learning optimized dense fault dynamics data, Journal of Marine Science and Engineering, 12 (2), DOI: 10.3390/jmse12020246.
• 19. Kanamori H., Anderson D.L., 1975, Theoretical basis of some empirical relations in seismology, Bulletin of Seismological Society of America, 65 (5), 1073-1095, DOI: 10.1785/BSSA0650051073.
• 20. Karczmarek P., Kiersztyn A., Pedrycz W., Al E., 2020, K-Means-based isolation forest, Knowledge- Based Systems, 195, DOI: 10.1016/j.knosys.2020.105659.
• 21. Kubo H., Naoi M., Kano M., 2024, Recent advances in earthquake seismology using machine learning, Earth, Planets and Space, 76, DOI: 10.1186/s40623-024-01982-0.
• 22. Liu F.T., Ting K.M., Zhou Z.H., 2008, Isolation Forest, Eighth IEEE International Conference on Data Mining, 413-422, DOI: 10.1109/ICDM.2008.17.
• 23. Liu F.T., Ting K.M., Zhou Z.-H., 2012, Isolation-based anomaly detection, ACM Transactions on Knowledge Discovery from Data, 6 (1), 1-39, DOI: 10.1145/2133360.2133363.
• 24. McGarr A., 1999, On relating apparent stress to the stress causing earthquake fault slip, Journal of Geophysical Research: Solid Earth, 104 (B2), 3003-3011, DOI: 10.1029/1998JB900083.
• 25. Melo A., Li Y., 2021, Geology differentiation by applying unsupervised machine learning to multiple independent geophysical inversions, Geophysical Journal International, 227 (3), 2058-2078, DOI: 10.1093/gji/ggab316.
• 26. Mollon G., Aubry J., Schubnel A., 2023, Laboratory earthquakes simulations - typical events, fault damage, and gouge production, Journal of Geophysical Research: Solid Earth, 128 (2), e2022JB025429, DOI: 10.1029/2022JB025429.
• 27. Moustafa S.S.R., Yassien M.H., Metwaly M., Faried A.M., Elsaka B.A., 2024, A unified seismicity catalog development for Saudi Arabia: multi-network fusion and machine learning-based anomaly, detection, Applied Science, 14 (16), DOI: 10.3390/app14167070.
• 28. O'Sullivan C., 2011, Particulate Discrete Element Modelling: A Geomechanics Perspective, Spon Press/Taylor and Francis, London, 576 pp.
• 29. Patel A.A., 2019, Hands-On Unsupervised Learning Using Python: How to Build Applied Machine Learning Solutions from Unlabeled Data, O'Reilly Media, 359 pp.
• 30. Prieto G.A., Shearer P.M., Vernon F.L., Kilb D., 2004, Earthquake source scaling and self-similarity estimation from stacking p and s spectra, Journal of Geophysical Research: Solid Earth, 109 (B8), DOI: 10.1029/2004JB003084.
• 31. Raschka S., Mirjalili V., 2019, Python Machine Learning: Machine Learning and Deep Learning with Python, scikit-learn, and TensorFlow 2, Packt Publishing, Third Edition, 772 pp.
• 32. Ren C.X., Dorostkar O., Rouet-Leduc B., Hulbert C., Strebel D., Guyer R.A., Johnson P.A., Carme- liet J., 2019, Machine learning reveals the state of intermittent frictional dynamics in a sheared granular fault, Geophysical Research Letters, 46 (13), 7395-7403, DOI: 10.1029/2019GL082706. Riviere
• 33. J., Lv Z., Johnson P.A., Marone C., 2018, Evolution of b-value during the seismic cycle: insights from laboratory experiments on simulated faults, Earth and Planetary Science Letters, 482, 407-413, DOI: 10.1016/j.epsl.2017.11.036.
• 34. Rouet-Leduc B., Hulbert C., Lubbers N., Barros K., Humphreys C.J., Johnson P.A., 2017, Machine learning predicts laboratory earthquakes, Geophysical Research Letters, 44 (18), 9276-9282, DOI: 10.1002/2017GL074677.
• 35. Senatorski P., 2023, Sejsmiczność, trzęsienia ziemi i lingiwstyka, Przegląd Geofizyczny, 68 (3-4), 185-197, DOI: 10.32045/PG-2023-042.
• 36. Seo J., Kim Y., Ha J., Kwak D., Ko M., Yoo M., 2024, Unsupervised anomaly detection for earthquake detection on Korea high-speed trains using autoencoder-based deep learning models, Scientific Reports, 14, DOI: 10.1038/s41598-024-51354-7.
• 37. Strumbelj E., Kononenko I., 2014, Explaining prediction models and individual predictions with feature contributions, Knowledge and Information Systems, 41 (3), 647-665, DOI: 10.1007/s10115- 013-0679-x.
• 38. Wang J., Cao J., Liu Z., 2024, Unsupervised machine learning-based multi-attributes fusion dim spot subtle sandstone reservoirs identification utilizing isolation forest, Geoenergy Science and Engineering, 234, DOI: 10.1016/j.geoen.2023.212626.
• 39. Weatherley D.K., Boros V.E., Hancock W.R., Abe S., 2010, Scaling benchmark of ESyS-Particle for elastic wave propagation simulations, 2010 IEEE Sixth International Conference E-Science, Brisbane, Australia, 277-283, DOI: 10.1109/eScience.2010.40.
• 40. Zhang K., Kang X., Li S., 2019, Isolation Forest for anomaly detection in hyperspectral images, IGARSS 2019, 2019 IEEE International Geoscience and Remote Sensing Symposium, Yokohama, Japan, 437-440, DOI: 10.1109/IGARSS.2019.8899812.

Uwagi

Opracowanie rekordu ze środków MNiSW, umowa nr POPUL/SP/0154/2024/02 w ramach programu "Społeczna odpowiedzialność nauki II" - moduł: Popularyzacja nauki (2025).