Machine learning strategies for lithostratigraphic classification based on geochemical sampling data: A case study in area of Chahanwusu River, Qinghai Province, China

Abstract

Based on the complex correlation between the geochemical element distribution patterns at the surface and the types of bedrock and the powerful capabilities in capturing subtle of machine learning algorithms, four machine learning algorithms, namely, decision tree (DT), random forest (RF), XGBoost (XGB), and LightGBM (LGBM), were implemented for the lithostratigraphic classification and lithostratigraphic prediction of a quaternary coverage area based on stream sediment geochemical sampling data in the Chahanwusu River of Dulan County, Qinghai Province, China. The local Moran’s I to represent the features of spatial autocorrelations, and terrain factors to represent the features of surface geological processes, were calculated as additional features. The accuracy, precision, recall, and F1 scores were chosen as the evaluation indices and Voronoi diagrams were applied for visualization. The results indicate that XGB and LGBM models both performed well. They not only obtained relatively satisfactory classification performance but also predicted lithostratigraphic types of the Quaternary coverage area that are essentially consistent with their neighborhoods which have the known types. It is feasible to classify the lithostratigraphic types through the concentrations of geochemical elements in the sediments, and the XGB and LGBM algorithms are recommended for lithostratigraphic classification.

Keywords

machine learning / geochemical sampling / lithostratigraphic classification / lithostratigraphic prediction / bedrock

References

[1] TolleK M, TansleyD S W, HeyA J G. the fourth paradigm: Data-intensive scientific discovery [J]. Proceedings of the IEEE, 2011, 99(8): 1334-1337

[2] ReichsteinM, Camps-VallsG, StevensB, JungM, DenzlerJ, CarvalhaisN, Prabhat. Deep learning and process understanding for data-driven earth system science [J]. Nature, 2019, 566(7743): 195-204

[3] BISHOP C M. Pattern recognition and machine learning [M]. Springer, 2007. https://www.springer.com/us/book/9780387310732.

[4] LakeB M, SalakhutdinovR, TenenbaumJ B. Human-level concept learning through probabilistic program induction [J]. Science, 2015, 350(6266): 1332-1338

[5] MOHRI M, ROSTAMIZADEH A, TALWALKAR A. Foundations of machine learning [M]. MIT Press, 2018. https://ieeexplore.ieee.org/document/6282245?reload=true&tp=&arnumber=6282245.

[6] DevriesP M R, ViegasF, WattenbergM, MeadeB J. Deep learning of aftershock patterns following large earthquakes [J]. Nature, 2018, 560(7720): 632-634

[7] RahmatiO, GolkarianA, BiggsT, KeesstraS, MohammadiF, DaliakopoulosI N. Land subsidence hazard modeling: Machine learning to identify predictors and the role of human activities [J]. J Environ Manage, 2019, 236: 466-480

[8] LiT-w, ShenH-f, YuanQ-q, ZhangX-c, ZhangL-pei. Estimating ground-level PM2.5 by fusing satellite and station observations: A geo-intelligent deep learning approach [J]. Geophysical Research Letters, 2017, 44(23): 985-993

[9] WangR, LiQ-y, YuH-m, ChenZ-c, ZhangY-j, ZhangL, CuiH-x, ZhangK. A category-based calibration approach with fault tolerance for air monitoring sensors [J]. IEEE Sensors Journal, 2020, 20(18): 10756-10765

[10] AdeliA, EmeryX, DowdP. Geological modelling and validation of geological interpretations via simulation and classification of quantitative covariates [J]. Minerals, 2017, 8(7): 8010007

[11] GonçalvesÍ G, KumairaS, GuadagninF. A machine learning approach to the potential-field method for implicit modeling of geological structures [J]. Computers and Geosciences, 2017, 103: 173-182

[12] MckayG, HarrisJ R. Comparison of the data-driven random forests model and a knowledge-driven method for mineral prospectivity mapping: A case study for gold deposits around the Huritz group and Nueltin suite, Nunavut, Canada [J]. Natural Resources Research, 2015, 25(2): 125-143

[13] Rodriguez-GalianoV, Sanchez-CastilloM, Chica-OlmoM, Chica-RivasM. Machine learning predictive models for mineral prospectivity: An evaluation of neural networks, random forest, regression trees and support vector machines [J]. Ore Geology Reviews, 2015, 71804-818

[14] ChenY-l, WuW. Isolation forest as an alternative data-driven mineral prospectivity mapping method with a higher data-processing efficiency [J]. Natural Resources Research, 2018, 28(1): 31-46

[15] ZhangN-n, ZhouK-f, LiD. Back-propagation neural network and support vector machines for gold mineral prospectivity mapping in the Hatu region, Xinjiang, China [J]. Earth Science Informatics, 2018, 11(4): 553-566

[16] SunT, ChenF, ZhongL-x, LiuW-m, WangY. GIS-based mineral prospectivity mapping using machine learning methods: A case study from tongling ore district, eastern China [J]. Ore Geology Reviews, 2019, 109: 26-49

[17] Keykhay-HosseinpoorM, KohsaryA H, Hossein-MorshedyA, PorwalA. A machine learning-based approach to exploration targeting of porphyry Cu-Au deposits in the Dehsalm district, eastern Iran [J]. Ore Geology Reviews, 2020, 116: 103234

[18] SunT, LiH, WuK-x, ChenF, ZhuZ, HuZ-juan. Data-driven predictive modelling of mineral prospectivity using machine learning and deep learning methods: A case study from southern Jiangxi province, China [J]. Minerals, 2020, 10(2): 10020102

[19] WangF-y, MaoX-c, DengH, ZhangB-Y. Manganese potential mapping in western Guangxisoutheastern Yunnan (China) via spatial analysis and modal-adaptive prospectivity modeling [J]. Transactions of Nonferrous Metals Society of China, 2020, 3041058-1070

[20] ZuoR-g, XiongY-hui. Big data analytics of identifying geochemical anomalies supported by machine learning methods [J]. Natural Resources Research, 2017, 27(1): 5-13

[21] ChenY-l, WuW. Application of one-class support vector machine to quickly identify multivariate anomalies from geochemical exploration data [J]. Geochemistry: Exploration, Environment, Analysis, 2017, 17(3): 231-238

[22] ChenL-r, GuanQ-f, FengB, YueH-q, WangJ-y, ZhangF. A multi-convolutional autoencoder approach to multivariate geochemical anomaly recognition [J]. Minerals, 2019, 9(5): 9050270

[23] WangZ-y, ZuoR-g, DongY-ni. Mapping geochemical anomalies through integrating random forest and metric learning methods [J]. Natural Resources Research, 2019, 2841285-1298

[24] GhezelbashR, MaghsoudiA, CarranzaE J M. Optimization of geochemical anomaly detection using a novel genetic K-means clustering (GKMC) algorithm [J]. Computers and Geosciences, 2020, 134104335

[25] WuR-y, ChenJ-l, ZhaoJ-n, ChenJ-d, ChenS-yu. Identifying geochemical anomalies associated with gold mineralization using factor analysis and spectrum-area multifractal model in Laowan district, Qinling-Dabie metallogenic belt, central China [J]. Minerals, 2020, 10(3): 10030229

[26] GuoZ-w, LaiJ-q, ZhangK-n, MaoX-c, LiuJ-xin. Geosciences in central south university: A state-of-the-art review [J]. Journal of Central South University, 2020, 274975-996

[27] SunJ, LiQ, ChenM-q, RenL, HuangG-h, LiC-y, ZhangZ-xuan. Optimization of models for a rapid identification of lithology while drilling-A win-win strategy based on machine learning [J]. Journal of Petroleum Science and Engineering, 2019, 176321-341

[28] YuL, PorwalA, HoldenE J, DentithM C. Towards automatic lithological classification from remote sensing data using support vector machines [J]. Computers and Geosciences, 2012, 45: 229-239

[29] PARAKH K, THAKUR S, CHUDASAMA B, TIRODKAR S, BHATTACHARYA A. Machine learning and spectral techniques for lithological classification [C]// SPIE Asia-Pacific Remote Sensing, 2016: 1–12. DOI: https://doi.org/10.1117/12.2223638.

[30] CracknellM J, ReadingA M. Geological mapping using remote sensing data: A comparison of five machine learning algorithms, their response to variations in the spatial distribution of training data and the use of explicit spatial information [J]. Computers and Geosciences, 2014, 63: 22-33

[31] Ordóñez-CalderÓnJ C, GelcichS. Machine learning strategies for classification and prediction of alteration facies: Examples from the Rosemont Cu-Mo-Ag skarn deposit, SE Tucson Arizona [J]. Journal of Geochemical Exploration, 2018, 194: 167-188

[32] DevA D, EdenM R. Formation lithology classification using scalable gradient boosted decision trees [J]. Computers and Chemical Engineering, 2019, 128: 392-404

[33] KitzigM, KepicA, GrantA. Near real-time classification of iron ore lithology by applying fuzzy inference systems to petrophysical downhole data [J]. Minerals, 2018, 8(7): 8070276

[34] XieY-x, ZhuC-y, ZhouW, LiZ-d, LiuX, TuM. Evaluation of machine learning methods for formation lithology identification: A comparison of tuning processes and model performances [J]. Journal of Petroleum Science and Engineering, 2018, 160: 182-193

[35] SunJ, LiQ, ChenM-q, RenL, HuangG-h, LiC-y, ZhangZ-xuan. Optimization of models for a rapid identification of lithology while drilling-A win-win strategy based on machine learning [J]. Journal of Petroleum Science and Engineering, 2019, 176321-341

[36] Savu-KrohnC, RantitschG, AuerP, MelcherF, GraupnerT. Geochemical fingerprinting of Coltan ores by machine learning on Uneven datasets [J]. Natural Resources Research, 2011, 20(3): 177-191

[37] CatéA, SchetselaarE, Mercier-LangevinP, RossP-S. Classification of lithostratigraphic and alteration units from drillhole lithogeochemical data using machine learning: A case study from the Lalor volcanogenic massive sulphide deposit, Snow Lake, Manitoba, Canada [J]. Journal of Geochemical Exploration, 2018, 188: 216-228

[38] HarrisJ R, GrunskyE C. Predictive lithological mapping of Canada’s north using random forest classification applied to geophysical and geochemical data [J]. Computers and Geosciences, 2015, 80: 9-25

[39] CostaI, TavaresF, OliveiraJ. Predictive lithological mapping through machine learning methods: A case study in the Cinzento Lineament, Carajás province, Brazil [J]. Journal of the Geological Survey of Brazil, 2019, 2(1): 26-36

[40] EllisD V, SingerJ MWell logging for earth scientists [M], 2007, Netherlands, Springer

[41] ChengQ-ming. Singularity theory and methods for mapping geochemical anomalies caused by buried sources and for predicting undiscovered mineral deposits in covered areas [J]. Journal of Geochemical Exploration, 2012, 122: 55-70

[42] ShahiH, GhavamiR, RouhaniA K. Detection of deep and blind mineral deposits using new proposed frequency coefficients method in frequency domain of geochemical data [J]. Journal of Geochemical Exploration, 2016, 162: 29-39

[43] ChenQ, JiaC-p, WeiJ-c, DongF-y, YangW-g, HaoD-c, JiaZ-w, JiY-han. Geochemical process of groundwater fluoride evolution along global coastal plains: Evidence from the comparison in seawater intrusion area and soil salinization area [J]. Chemical Geology, 2020, 552: 119779

[44] CHEN Qiao, HAO De-cheng, WEI Jiu-chuan, JIA Cui-ping, WANG Hong-mei, SHI Long-qing, LIU Song-liang, NING Fang-zhu, AN Mao-guo, JIA Zhi-wei, DONG Fang-ying, JI Yu-han. The influence of high-fluorine groundwater on surface soil fluorine levels and their FTIR characteristics [J]. Arabian Journal of Geosciences, 2020, 13: No. 383. DOI: https://doi.org/10.1007/s12517-020-05346-2.

[45] AnandR R, AspandiarM F, NobleR R P. A review of metal transfer mechanisms through transported cover with emphasis on the vadose zone within the Australian regolith [J]. Ore Geology Reviews, 2016, 73(3): 394-416

[46] ZaremotlaghS, HezarkhaniA. The use of decision tree induction and artificial neural networks for recognizing the geochemical distribution patterns of LREE in the Choghart deposit, central Iran [J]. Journal of African Earth Sciences, 2016, 12837-46

[47] ZhangB-y, ChenY-r, HuangA-s, LuH, ChengQ-ming. Geochemical field and its roles on the 3D prediction fo concealed ore-bodies [J]. Acta Petrologica Sinica, 2018, 34(2): 352-362(in Chinese)

[48] WangL-f, WuX-b, ZhangB-y, LiX, HuangA-s, MengF, DaiP-yao. Recognition of significant surface soil geochemical anomalies via weighted 3d shortest-distance field of subsurface orebodies: A case study in the Hongtoushan copper mine, NE China [J]. Natural Resources Research, 2019, 28(3): 587-607

[49] CortesC, VapnikV. Support-vector networks [J]. Machine Learning, 1995, 20(3): 273-297

[50] QuinlanJ R. Induction of decision trees [J]. Machine Learning, 1986, 1(1): 81-106

[51] KodikaraJ R L, WoldaiT. Spectral indices derived, non-parametric decision tree classification approach to lithological mapping in the Lake Magadi area, Kenya [J]. International Journal of Digital Earth, 2017, 11(10): 1020-1038

[52] BreimanL. Random forests [J]. Machine Learning, 2001, 45(1): 5-32

[53] FREUND Y, SCHAPIRE R E. Experiments with a new boosting algorithm [C]// International Conference on Machine Learning: Proceedings of the Thirteenth International Conference. 1996: 148–156. http://cseweb.ucsd.edu/~yfreund/papers/boostingexperiments.pdf.

[54] FriedmanJ H. Stochastic gradient boosting [J]. Computational Statistics & Data Analysis, 2002, 38(4): 367-378

[55] ChenT-q, GuestrinC. XGBoost: A scalable tree boosting system [C]. the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, San Francisco, California, USA, Association for Computing Machinery, 785794

[56] KeG-l, MengQ, FinleyT, WangT-f, ChenW, MaW-d, YeQ-w, LiuT-Yan. LightGBM: A highly efficient gradient boosting decision tree [C]. Neural Information Processing Systems 30 (NIPS 2017), 2017, Long Beach, CA, USA, Neural Information Processing Systems Conference, 31493157

[57] JiangK, WangS-d, HuY-j, PuS-z, DuanH, WangZ-wen. Lithology identification model by well logging based on boosting tree algorithm [J]. Well Logging Technology, 2018, 42(4): 29-34(in Chinese)

[58] XIE Zheng-wen, ZHU Chen-yang, LU Yue, ZHU Zheng-wei. Towards optimization of boosting models for formation lithology identification [J]. Mathematical Problems in Engineering, 2019, 5309852. DOI: https://doi.org/10.1155/2019/5309852.

[59] Asante-OkyereS, ShenC-b, ZiggahY Y, RulegeyaM M, ZhuX-feng. A novel hybrid technique of integrating gradient-boosted machine and clustering algorithms for lithology classification [J]. Natural Resources Research, 2019, 29(4): 2257-2273

[60] GOODFELLOWI J, POUGET-ABADIE J, MIRZA M, XU Bing, WARDE-FARLEY D, OZAIR S, COURVILLE A, BENGIO Y. Generative adversarial networks [C]// Advances in Neural Information Processing Systems. Montréal, Canada: 2014: 2672–2680. https://arxiv.org/pdf/1406.2661v1.pdf.

[61] XuS-t, ZhouY-zhang. Artificial intelligence identification of ore minerals under microscope based on deep learning algorithm [J]. Acta Petrologica Sinica, 2018, 34(11): 3244-3252(in Chinese)

[62] LiG-h, QiaoY-h, ZhengY-f, LiY, WuW-jiang. Semi-supervised learning based on generative adversarial network and its applied to lithology recognition [J]. IEEE Access, 2019, 7: 67428-67437

[63] LiuC-z, LiM-c, ZhangY, HanS, ZhuY-qin. An enhanced rock mineral recognition method integrating a deep learning model and clustering algorithm [J]. Minerals, 2019, 9516

[64] AnselinL. Local indicators of spatial association—LISA [J]. Geographical Analysis, 1995, 27(2): 93-115

[65] QuinlanJ RC4.5: Programs for machine learning [M], 1993, San Francisco, Morgan Kaufmann Publishers Inc

[66] BreimanL, FriedmanJ, OlshenR A, StoneC JClassification and regression trees [M], 1984, New York, Chapman and Hall