Application of machine learning methods to classify quartzites by chemical composition: the influence of trace elements and geochemical identification

   Research subject. Quartzites from Different Sites of the East Sayan Quartz-Bearing Region.

   Aim. The aim of this study is to apply machine learning methods to effectively classify quartzite samples by their chemical composition, including the identification of key trace elements such as manganese and the detection of geochemical differences between samples.

   Materials and methods. The study used chemical analysis data from 776 quartzite samples, which were interpreted using machine learning methods. Standard data preprocessing techniques such as normalization were applied as well as data augmentation using SMOTE to solve the class imbalance problem. As a result, the CatBoost algorithm was selected, which showed high classification accuracy.

   Results. Cross-validation results showed that the CatBoost algorithm achieved classification accuracy of up to 97 %. The importance of features indicates that manganese is a key element in the classification of samples, while elements such as aluminum and potassium have a supporting effect. The analysis of the classification by color of quartzites with an accuracy of 0.94 was also successfully carried out.

   Conclusions. The study demonstrates the effectiveness of applying machine learning methods to the analysis of the chemical composition of quartzites, providing new opportunities for geochemical and archaeological research.

1. Ayurzhanayeva D.Ts., Fedorov A.M., Mazukabzov A.M., Nepomnyaschikh A.I., Ochirova E.A., Posokhov V.F. (2020) Mechanisms of High-Purity Quartzite Formation at the Bural-Sardyk Deposit (Russia). Geol. Geophys., 61(10), 1316-1330. doi: 10.15372/GiG2023128

2. Baak M., Koopman R., Snoek H., Klous S. (2020) A new correlation coefficient between categorical, ordinal and interval variables with Pearson characteristics. Comput. Stat. Data Anal., 152, 107043. doi: 10.1016/j.csda.2020.107043

3. Bajpai J., Singh S. (2019) On orthogonal hypergeometric groups of degree five. Transact. Amer. Math. Soc., 372(11), 7541. doi: 10.1090/tran/7677

4. Breiman L. (2001) Random Forests. Machine Learning, 45(1), 5-32. doi: 10.1023/A:1010933404324

5. Chawla N.V., Bowyer K.W., Hall L.O., Kegelmeyer W.P. (2002) SMOTE: Synthetic minority over-sampling technique. J. Artific. Intellig. Res., 321-357. doi: 10.1613/jair.953

6. Fedorov A.M., Makrygina V.A., Mazukabzov A.M., Nepomnyashchikh A.I., Ayurzhanayeva D.Ts., Volkova M.G. (2021) Resources of quartz raw materials, Gargan block, East Sayan quartzite-bearing area. Georesources, 23(4), 96-106. doi: 10.18599/grs.2021.4.11

7. Fedorov A.M., Makrygina V.A., Nepomnyaschikh A.I., Zhaboedov A.P., Parshin A.V., Posokhov V.F., Sokolnikova Yu.V. (2019) Geochemistry and petrology of superpure quartzites from East Sayan Mountains, Russia. Acta Geochim. 38(1), 22-39. doi: 10.1007/s11631-018-0268-5

8. Götze J., Möckel R. (2012) Quartz: Deposits, Mineralogy and Analytics. Spring. Geol., Ó Springer-Verlag Berlin Heidelberg.

9. Hnila P., Frahm E., Gilibert A. et al. (2025) “Open Sourcing” Workflow and Machine Learning Approaches for Attributing Obsidian Artifacts to Their Volcanic Origins: A Feasibility Study from the South Caucasus. J. Archaeol. Method. Theory, 32(28).

10. Ke G., Meng Q., Finley Th., Wang T., Chen W., Ma W., Ye Q., Liu T. (2017) LightGBM: A Highly Efficient Gradient Boosting Decision Tree. Proceedings of the 31sup>st</sup> Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, 4-9 December 2017, 314s9-3157. https://dl.acm.org/ doi: 10.5555/3294996.3295074

11. Lundberg S.M., Lee S.-I. (2017) A Unified Approach to Interpreting Model Predictions. Advances in Neural Information Processing Systems, 30.

12. Manning C.D., Raghavan P., Schütze H. (2008) Introduction to Information Retrieval. Cambridge University Press, 234-265.

13. Müller A., Wanvik J.E., Ihlen P.M. (2012) Petrological and Chemical Characterisation of High-Purity Quartz Depo sits with Examples from Norway. (Eds J. Götze, R. Möckel). Quartz: Deposits, Mineralogy and Analytics. Springer Geology. Springer, Berlin, Heidelberg.

14. Nepomnyashchikh A.I., Fedorov A.M., Zhaboev A.P., Volkova M.G. (2023) High-Purity Quartzite from East Sayan. Geol. Geophys., 64(8), 1205-1215. doi: 10.15372/GiG2023128

15. Pitblado B.L., Cannon M.B., Neff H., Dehler C.M., Nelson S.T. (2013) LA-ICP-MS analysis of quartzite from the Upper Gunnison Basin, Colorado. J. Archaeol. Sci., 40(4), 2196-2216. doi: 10.1016/j.jas.2012.11.016

16. Pitblado B.L., Dehler C., Neff H., Nelson S.T. (2008) Pilot study experiments sourcing quartzite, Gunnison Basin, Colorado. Geoarchaeology, 23(6), 742-778. doi: 10.1002/gea.20240

17. Prokhorenkova L., Gusev G., Vorobev A., Dorogush A.V., Gulin A. (2017) CatBoost: Unbiased boosting with cate gorical features. arXiv:1706.09516v5. doi: 10.48550/arXiv.1706.09516

18. Shah S.A., Shao Y., Zhang Y., Zhao H., Zhao L. (2022) Texture and Trace Element Geochemistry of Quartz : A Review. Minerals, 12, 1042. doi: 10.3390/min12081042

19. van der Maaten L.J.P. (2014) Accelerating t-SNE using Tree-Based Algorithms. J. Mach. Learn. Res., 15(93), 3221-3245.

20. Wang Y., Qiu K.-F., Müller A., Hou Z.-L., Zhu Z.-H., Yu H.-C. (2021) Machine learning prediction of quartz forming-environments. J. Geophys. Res.: Solid Earth, 126, e2021JB021925. doi: 10.1029/2021JB021925

21. Yu H.F., Huang F.L., Lin C.J. (2011) Dual coordinate descent methods for logistic regression and maximum entropy models. Mach. Learn., 85, 41-75. doi: /10.1007/s10994-010-5221-8

22. Yudovich Ya.E., Ketris M.P. (2013) Manganese geochemistry in hypergenesis processes : A review. Biosphere, 5(1), 21-36.

23. Zhu G.-D., Niu Y.-Y., Liao S.-B., Ruan L., Zhang X.-H. (2023) Discrimination of Quartz Genesis Based on Explainable Machine Learning. Minerals, 13, 997. doi: 10.3390/min13080997