CLUSTERING ALGORITHMS IN EXPRESS-ANALYSIS OF SEISMIC DATA

A grand problem in continuous geomonitoring is constituted by a huge amount of accumulated data to be analyzed and interpreted. In this regard, it is relevant to carry out the preliminary express-analyses of data in order to select representative points to be in spotlight in the first place in the studied rock mass. By way of approach to this problem solution, the authors propose the clustering analysis which produces clustering centers usable as representative points. The scope of the discussion embraces three clustering algorithms: k-means, Mean Shift and DBSCAN. Their efficiency and suitability to the seismic data analysis is assessed by comparing the results of the algorithms with the representative points found by expert for the pre-assigned set of data. The quality of the clustering procedure is evaluated by the Calinski–Harabasz and Davies–Bouldin indexes, as well as the silhouette coefficient. The set of the seismic data was composed of numerical values of stress state (rock pressure) and fluid flow potential in rock mass obtained in local prediction. The obtained results allow concluding that the best clustering algorithm is DBSCAN, and it is applicable to preliminary express-analysis of seismic data.


For citation: Abdrakhmanov M. I., Lapin S. E., Shnayder I. V. Clustering algorithms in express-analysis of seismic data. MIAB. Mining Inf. Anal. Bull. 2019;(6):27-44. [In Russ]. DOI: 10.25018/0236-1493-2019-06-0-27-44.

Keywords

Geomonitoring, seismic data analysis, clustering analysis, express-analysis, clustering metrics.

Issue number: 6
Year: 2019
ISBN: 0236-1493
UDK: 622.817.4
DOI: 10.25018/0236-1493-2019-06-0-27-44
Authors: Abdrakhmanov M. I., Lapin S. E., Shnayder I. V.

About authors: M.I. Abdrakhmanov, Cand. Sci. (Eng.), Chief Specialist, e-mail: marat-ab@mail.ru, LLC «Information mining technologies», Ekaterinburg, Russia, S.E. Lapin (1), Cand. Sci. (Eng.), Senior Researcher, e-mail: sergei.l@bk.ru, I.V. Shnayder (1), Graduate Student, 1) Ural State Mining University, 620144, Ekaterinburg, Russia. Corresponding author: M.I. Abdrakhmanov, e-mail: marat-ab@mail.ru.

REFERENCES:

1. Prikaz Rostekhnadzora ot 19.11.2013 № 550 «Ob utverzhdenii Federal'nykh norm i pravil v oblasti promyshlennoy bezopasnosti «Pravila bezopasnosti v ugol'nykh shakhtakh» [Approval of Federal Code on Industrial Safety: Safety Regulations for Coal Mines. Rostekhnadzor Order No. 550 dated November 19, 2013]. [In Russ].

2. Official site OOO «INGORTEH», http://ingortech.ru/produktsiya/statsionarnye-sistemy/paragraf-41-pb/kontrol-gornogo-massiva-p-41-pb.

3. Patent US6498989, https://patents.google.com/patent/US6498989.

4. Lapin E. S., Pisetskiy V. B., Babenko A. G., Patrushev Yu. V. Mikon-GEO—on-line detection and monitoring of hazardous geo-gas-dynamic event initiation and growth in underground mineral mining. Bezopasnost' truda v promyshlennosti. 2012, no 4, pp. 18—22. [In Russ].

5. Calinski T., Harabasz J. A dendrite method for cluster analysis. Communications in Statistics. 1974, vol. 3, pp. 1—27.

6. Davies D. L., Bouldin D. W. A Cluster Separation Measure. IEEE Transactions on Pattern Analysis and Machine Intelligence. PAMI-1. 1979, pp. 224—227.

7. Rousseeuw P. J. Silhouettes: a Graphical Aid to the Interpretation and Validation of Cluster Analysis. Computational and Applied Mathematics 20. 1987, pp. 53—65.

8. Lloyd S. Least square quantization in PCM’s. IEEE Transactions on Information Theory. vol. 28, pp. 129—137.

9. Comaniciu D., Meer P. Mean shift. A robust approach toward feature space analysis. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002.

10. Ester M., Kriegel H. P., Sander J., Xu X. A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise. Proceedings of the 2nd International Conference on Knowledge Discovery and Data Mining, Portland, OR, AAAI Press., 1996, pp. 226—231.

11. Sebastian Raschka. Python Machine Learning, 1st Edition. Packt Publishing Ltd. 2015, 454 p.

12. Arthur D., Vassilvitskii S. k-means++: The advantages of careful seeding. Proceedings of the eighteenth annual ACM-SIAM symposium on Discrete algorithms, Society for Industrial and Applied Mathematics. 2007, pp. 1027—1035.

13. Sander J., Ester M., Kriegel H. P., Xu X. Density-Based Clustering in Spatial Databases: The Algorithm GDBSCAN and Its Applications. Data Mining and Knowledge Discovery. Berlin: Springer-Verlag. 1998 (2), pp. 169—194.

14. Brendan J. F., Delbert D. Clustering by passing messages between data points. Science. 2007. No 15, pp. 972—979.

15. Ng A., Jordan M., Weiss Y. On spectral clustering: analysis and an algorithm. NIPS'01 Proceedings of the 14th International Conference on Neural Information Processing Systems: Natural and Synthetic. 2001, pp. 849—856.

16. Amorim R. C., Hennig C. Recovering the number of clusters in data sets with noise features using feature rescaling factors. Information Sciences. 324. 2015. 126—145.

17. Hamerly G., Drake J. Accelerating Lloyd's algorithm for k-means clustering. PartitionalClustering Algorithms. 2015, pp. 41—78.

18. Campello R. J. G. B., Moulavi D, Zimek A., Sander J. Hierarchical Density Estimates for Data Clustering, Visualization, and Outlier Detection. ACM Transactions on Knowledge Discovery from Data. 2015. vol 10, pp. 5:1—5:51.

19. Pisetskiy V. B., Lapin S. E., Zudilin A. E., Patrushev Yu. V., Shnayder I. V. Procedure and commercial application results of Mikon-GEO seismic monitoring system in underground mining of ore and coal deposits. Problemy nedropol'zovaniya. 2016, pp. 58—64. [In Russ].

20. Pisetskiy V. B., Robert Huang, Patrushev Yu. V., Zudilin A. E., Shnayder I. V., Shirobokov M. P. Test data of seismic monitoring systems for rock mass stability in construction of highway tunnels in China. Dobyvayushchaya promyshlennost'. 2017, no 2 (06), pp. 108. [In Russ].

21. Pisetskiy V. B., Vlasov V. V., CHerepanov V. P., Abaturova I. V., Zudilin A. E., Patrushev Yu. V., Aleksandrova A. V. Rock mass stability prediction based on seismic location method in underground construction. Inzhenernye izyskaniya. 2014, no 9—10, pp. 46—51. [In Russ].

22. Yakovlev D. V., Lazarevich T. I., Polyakov A. N. Principles of constructing rock mass monitoring systems based on analysis of actual risks in underground mineral mining. Gornyy informatsionno-analiticheskiy byulleten’. 2015. Special edition 7, pp. 471—481. [In Russ].

Mining World Russia
Subscribe for our dispatch