Supervised learning for categorization of geological domain
DOI:
https://doi.org/10.11606/issn.2316-9095.v25-222751Keywords:
Machine Learning, Supervised Learning, Geological Model, Gold DepositAbstract
The geological model needs to be updated when new drill holes are added. This requires time and knowledge of the mineral deposit, as the new samples need to be categorized according to the relevant geological properties. This study employed six supervised machine learning algorithms (naive Bayes, k-Nearest Neighbors (kNN), Support Vector Machines (SVM), decision trees, random forest, and neural networks) to perform geological interpretation of a gold deposit of the metasedimentary type, located in the east-central region of the state of Bahia, with the objective of evaluating the ability of these algorithms to accurately classify the geological domains of new samples in a database. The results indicated that the random forest and neural networks algorithms successfully reproduced the geological interpretation of the mineral deposit, as the models generated were similar to those manually created by a geologist. This great performance is supported by the accuracies and precision obtained, which were 0.87 and 0.89, respectively. Therefore, it is recommended to use these algorithms for optimizing the geological model update process. However, the naive Bayes, kNN, SVM, and decision tree algorithms failed to categorize the geological domains of the samples correctly. As a result, some regions in the models exhibited distorted geological layers and a lack of statigraphic order due to interpretation erros. This is evidenced by the low accuracies and precision obtained-0.48, 0.73, and 0.75. Therefore, these algorithms are unsuitable for this task.
Downloads
References
Abzalov, M. (2016). Quantitative Geological Model. In: Abzalov, M (Eds). Applied mining Geology (v.1, 335-3350). Switzerland: Springer. https://doi.org/10.1007/978-3-319-39264-6
Armstrong, M. (1998). Regionalized Variables. In: Armstrong, M. Basic Linear Geostatistics (v.1, 20-34). Berlin, Heidelberg: Springer. https://doi.org/10.1007/978-3-642-58727-6
Awad, M., Khanna, R. (2015). Machine Learning. In: Awad, M. e Khanna, R. Efficient Learning Machines: Theories, Concepts, and Applications for Engineers and System Designers. Berkeley: Apress open, 1-18. https://doi.org/10.1007/978-1-4302-5990-9_1
Chawla, N. V., Japkowicz, N., Kotcz, A. (2004). Editorial: Special issue on learning from imbalanced data sets. ACM SIGKDD Explorations Newsletter, 6, 1-6. https://doi.org/10.1145/1007730.1007733
Cowan, E. J., Beatson, R. S., Ross, H. J., Fright, W. R., McLennan, T. J., Evans, T. R., Carr, J. C., Lane, R. G., (2003). Practical Implicit Geological Modelling. In: Dominy, S. (ed), 5th International Mining Geology Conference, 8, 89-99. Australia: The Australasian Institute of Mining and Metallurgy. Available at: https://www.researchgate.net/publication/281685127_Practical_Implicit_Geological_Modelling. Accessed on: Dec 16, 2025.
Dougherty, G. (2013). Classification. In: Dougherty, G (Ed). Patter Recognition and Classification- An Introduction, v.1, 9-26. New York: Springer. https://doi.org/:10.1007/978-1-4614-5323-9
Garayp, E., Frimmel, H. E. (2023). A modified paleoplacer model for the metaconglomerate-hosted gold deposits at Jacobina, Brazil. Mineralium Deposita. https://doi.org/10.1007/s00126-023-01220-9
Mao, X., Zhang, W., Liu, Z., Ren, J., Bayless, R. C., Deng, H. (2020). 3D Mineral Prospectivity Modeling for the Low-Sulfidation Epithermal Gold Deposit: A Case Study of the Axi Gold Deposit, Western Tianshan, NW China. Mineral, 10(3), 233. https://doi.org/10.3390/min10030233
Rasera, L. G. (2014). Geoestatística de múltiplos pontos aplicada à simulação de modelos geológicos em grids estratigráficos. Dissertação (Mestrado). Porto Alegre: Escola de Engenharia de Minas, Metalúrgica e de Materiais, Universidade de Rio Grande do Sul -UFRGS. Available at: https://lume.ufrgs.br/bitstream/10183/117142/2/000929816.pdf.txt5178f9de3f751d6f65932823152e5c1fMD52THUMBNAIL000929816.pdf.jpg000929816.pdf.jpgGenerated. Accessed on: Dec 16, 2025.
Rossi, M. E., Deutsch, C. V. (2014). Geological Controls and Block Modeling. In: Rossi, M.E. & Deutsch, C. V. Mineral Resource Estimation ( 29-50). New York: Springer. https://doi.org/10.1007/978-1-4020-5717-5_3
Samuel, A. L. (1959). Some studies in machine learning using the game of checkers. IBM Journal of Research and Development, v. 3, 210-229. https://doi.org/10.1147/rd.33.0210
Singh, D., Singh, B. (2020). Investigating the impact of data normalization on classification performance. Applied Soft Computing, 97, 105524. https://doi.org/10.1016/j.asoc.2019.105524
Zhang, S. E., Nwaila, G. T., Bourdeau, J. E., Ghorbani, Y., Carranza, E. J. M. (2023). Machine Learning-Based Delineation of Geodomain Boundaries: A Proof-of-Concept Study Using Data from the Witwatersrand Goldfields. Natural Resources Research, 32, 879-900. https://doi.org/10.1007/s11053-023-10159-7
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Marcela Martins, Marcelo Monteiro da Rocha, Camila Duelis Viana
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Authors who publish in this journal shall comply with the following terms:
- Authors keep their copyright and grant to Geologia USP: Série Científica the right of first publication, with the paper under the Creative Commons BY-NC-SA license (summary of the license: https://creativecommons.org/licenses/by-nc-sa/4.0 | full text of the license: https://creativecommons.org/licenses/by-nc-sa/4.0/legalcode) that allows the non-commercial sharing of the paper and granting the proper copyrights of the first publication in this journal.
- Authors are authorized to take additional contracts separately, for non-exclusive distribution of the version of the paper published in this journal (publish in institutional repository or as a book chapter), granting the proper copyrights of first publication in this journal.
- Authors are allowed and encouraged to publish and distribute their paper online (in institutional repositories or their personal page) at any point before or during the editorial process, since this can generate productive changes as well as increase the impact and citation of the published paper (See The effect of Open Access and downloads on citation impact).
