GEOPHYSICAL RESEARCH, 2021, vol. 22, no. 1, pp. 25-39. https://doi.org/10.21455/gr2021.1-2
UDC 004.891.3; 550.8.053
Abstract References Full text (in Russian)
NEURAL NETWORK CLUSTER ANALYSIS OF AREAL GEOPHYSICAL DATA OF THE KHIBINY-LOVOZERO VOLCANO-PLUTONIC ORE-BEARING COMPLEX (KOLA PENINSULA) BY SELF-ORGANIZING MAPS OF KOHONEN METHOD
I.I. Nikulin(1), A.A. Samsonov(2)
(1) LLC Norilskgeologiya, St. Petersburg, Russia
(2) Lomonosov Moscow State University, Moscow, Russia
Corresponding author: I.I. Nikulin (e-mail: iinikulin@gmail.com)
Abstract. The method of automatic express-interpretation of areal geophysical data is described using the example of the Khibiny-Lovozero volcano-plutonic complex. Within the framework of this method a complex geophysical-mathematical model was built, consisting of fifteen levels of hierarchy and based on the use of the mathematical apparatus of artificial neural networks. The method of self-organizing Kohonen maps was used in the processing of geophysical survey data. This is a mathematical apparatus of fuzzy logic, which artificial neural network is trained without a teacher. The formation of groups of clusters that characterize to the fullest extent the possible connections between multidimensional geophysical data is substantiated. As well as the presence of relationships between them is analyzed by identifying correlation dependences. The analysis of various geophysical transformants using self-organizing Kohonen maps is carried out. The clusters that as a result of the study reflect the picture of a major Paleozoic ore-magmatic system in the northeast of the Fennoscandian Shield were calculated using an artificial neural network. This system brings together the Khibiny and Lovozero plutons, Kurginskaya intrusion, volcanic formations and numerous swarms of alkaline dikes. A number of input indicators-representatives of their groups were determined and, on their basis, a geophysical-mathematical model in the form of a two-dimensional map of clusters using fuzzy logic tools was built. Terminological sets for each group of clusters were formed, the form of membership functions of previously unknown geological objects was specified according to the new interpreted data and their parameters in relation to the main ring of the Khibiny array, that controls rare-earth-titanium-aluminum-phosphorus deposits, and Fedoro-Panskaya Tundras (platinum-paladium deposits). Potential objects for geological exploration near the Lovozero eudialyte zircon rare-earth deposit are proposed. The conducted experimental study confirmed the adequacy of the constructed model and the effectiveness of its use for the purpose of express analysis of geophysical data and decision-making in geological prospecting tasks.
Keywords: gravity exploration, magnetic exploration, hierarchical model, artificial neural network, Kohonen map, fuzzy logic, cluster analysis.
Reference
Arzamastsev A.A., Arzamastseva L.V., Zhirova A.M., Glaznev V.N., Model of formation of the Khibiny-Lovozero ore-bearing volcanic-plutonic complex, Geology of Ore Deposits, 2013, vol. 55, no. 5, pp. 341-356.
Cavalcanti R.B.C., Pimentel B.A., de Almeida C.W.D., de Souza R.M.C.R., A Multivariate Fuzzy Kohonen Clustering Network, 2019 International Joint Conference on Neural Networks (IJCNN), 2019, https://ieeexplore.ieee.org/document/8852243. doi: 10.1109/IJCNN.2019.8852243
Diachenko V., Liashenko O., Ibrahim B.F., Mikhal O., Koltun Yu., Kohonen Network with Parallel Training: Operation Structure and Algorithm, International Journal of Advanced Trends in Computer Science and Engineering, 2019, vol. 8, no. 1, pp. 35-38. doi.org/10.30534/ijatcse/2019/0681.22019
Hu Z., Bodyanskiy Y.V., Tyshchenko O.K., Kohonen Maps and Their Ensembles for Fuzzy Clustering Tasks, Self-Learning and Adaptive Algorithms for Business Applications, Emerald Publishing Limited, 2019, pp. 51-77, doi.org/10.1108/978-1-83867-171-620191004
Ivanuk G.U., Gorayinov P.M., Pakhomovski Y.A., Konopleuva N.G., Yakovenchuk V.N., Bazai A.B., Kalashnikov A.O., Samoorganizatsiya rudnykh kompleksov. Sinergeticheskie printsipy prognozirovaniya i poiskov mestorozhdenii poleznykh iskopaemykh (Self-organization of ore complexes. Synergetic principles of forecasting and prospecting for mineral deposits), Moscow: GEOKART-GEOS, 2009, 392 p. [in Russian].
Kalashnikov A.O., Nikulin I.I., Stepenshchikov D.G., Unsupervised geochemical classification and automatic 3D mapping of the Bolshetroitskoe high-grade iron ore deposit (Belgorod Region, Russia), Sci Rep., 2020, vol. 10, no. 17861, 13 р. https://doi.org/10.1038/s41598-020-74505-y
Kirillov I.E., Morozov I.N., Oleynik A.G., Development of models based on neural networks and fuzzy logic for express analysis of ore-dressing processes, Trudy Kolskogo nauchnogo centra RAN (Proceedings of the Kola Scientific Center of the Russian Academy of Sciences), 2013, no. 5 (18), pp. 152-159. [in Russian].
Kohonen T., Samoorganizuyushchiesya karty. Seriya: Adaptivnye i intellektual'nye sistemy (Self-organizing cards. Series: Adaptive and intelligent systems), Moscow: Binom. Laboratoriya znanii, 2008, 656 p. [in Russian].
Larose D.T., Larose C.D., Discovering Knowledge in Data: An Introduction to Data Mining, John Wiley & Sons, 2014, 336 p. doi.org/10.1002/9781118874059
Nikulin I.I., Savko A.D., Iron ore weathering crust of the Belgorod region of the Kursk magnetic anomaly, Trudy nauchno-issledovatelskogo institute geologii Voronezhskogo gosudarstvennogo universiteta (Proceedings of the Research Institute of Geology, Voronezh State University), Voronezh: Voronezh State University, 2015, issue 85, 102 p. [in Russian].
Samanta S., Utkarsha S., Gour B., Tiwari Kr.V., Fast Character Recognition Using Kohonen Neural Network, 2018 International Conference on Advanced Computation and Telecommunication (ICACAT), 2018. https://ieeexplore.ieee.org/document/8933627. doi: 10.1109/ICACAT.2018.8933627
Samoorganizuuzchiiesaya karty Kohonena – matematicheskii apparat (Kohonen's self-organizing maps – mathematical apparatus), URL: https://basegroup.ru/community/articles/som. Accessed 29.09.2020. [in Russian].
Shablinsky G.N., To the question of the deep structure of the Khibiny and Lovozersky plutons, Trudy Leningradskogo obshchestva estestvoispytatelei (Proceedings of the Leningrad Society of Naturalists), 1963, vol. 74, pp. 41-43. [in Russian].
Silva I.N., Spatti D.H., Flauzino R.A., Liboni L.H.B., dos Reis Alves F.S., Self-Organizing Kohonen Networks, Artificial Neural Networks, Springer, Cham, 2017, pp. 157-172. doi.org/10.1007/978-3-319-43162-8_8
Todeva E., Knoke D., Keskinova D., Multi-stage clustering with complementary structural analysis of 2-mode networks, 2019 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 2019, pp. 771-778. doi.org/10.1145/3341161.3344781