Machine learning model comparison and ensemble for predicting different morphological fractions of heavy metal elements in tailings and mine waste

Abstract

Driven by rapid technological advancements and economic growth, mineral extraction and metal refining have increased dramatically, generating huge volumes of tailings and mine waste (TMWs). Investigating the morphological fractions of heavy metals and metalloids (HMMs) in TMWs is key to evaluating their leaching potential into the environment; however, traditional experiments are time-consuming and labor-intensive. In this study, 10 machine learning (ML) algorithms were used and compared for rapidly predicting the morphological fractions of HMMs in TMWs. A dataset comprising 2376 data points was used, with mineral composition, elemental properties, and total concentration used as inputs and concentration of morphological fraction used as output. After grid search optimization, the extra tree model performed the best, achieving coefficient of determination (R²) of 0.946 and 0.942 on the validation and test sets, respectively. Electronegativity was found to have the greatest impact on the morphological fraction. The models’ performance was enhanced by applying an ensemble method to the top three optimal ML models, including gradient boosting decision tree, extra trees and categorical boosting. Overall, the proposed framework can accurately predict the concentrations of different morphological fractions of HMMs in TMWs. This approach can minimize detection time, aid in the safe management and recovery of TMWs.

Keywords

tailings and mine waste / morphological fractions / model comparison / machine learning / model ensemble

References

[1] Li D-w, Ramos A O, Bah A, et al.. Valorization of lead-zinc mine tailing waste through geopolymerization: Synthesis, mechanical, and microstructural properties [J]. Journal of Environmental Management, 2024, 349: 119501

[2] Yildiz T D. Considering the development levels of countries, contributions of mineral recovery from mining tailings and urban mining wastes to sustainability criteria–A review [J]. Resources Policy, 2024, 99: 105399

[3] Calderon J L, Smith N M, Bazilian M D, et al.. Critical mineral demand estimates for low-carbon technologies: What do they tell us and how can they evolve? [J]. Renewable and Sustainable Energy Reviews, 2024, 189: 113938

[4] Qi C-c, Hu T, Zheng Y, et al.. Global and regional patterns of soil metal(loid) mobility and associated risks [J]. Nature Communications, 2025, 16(1): 2947

[5] Surehali S, Han T-h, Huang J, et al.. On the use of machine learning and data-transformation methods to predict hydration kinetics and strength of alkali-activated mine tailings-based binders [J]. Construction and Building Materials, 2024, 419: 135523

[6] Karacan C Ö, Erten O, Martín-Fernández J A. Assessment of resource potential from mine tailings using geostatistical modeling for compositions: A methodology and application to Katherine Mine site, Arizona, USA [J]. Journal of Geochemical Exploration, 2023, 245: 107142

[7] Lin L, Liu X, Liang Y, et al.. Analysis of mineral phases in heavy-metal hazardous waste under the interdisciplinary scope of data science and chemistry [J]. Progress in Chemistry, 2021, 33(122163-2172

[8] Piatak N M, Seal R R. Mineralogy and environmental geochemistry of historical iron slag, Hopewell Furnace National Historic Site, Pennsylvania, USA [J]. Applied Geochemistry, 2012, 27(3623-643

[9] Bolan N, Kunhikrishnan A, Thangarajan R, et al.. Remediation of heavy metal(loid)s contaminated soils-To mobilize or to immobilize? [J]. Journal of Hazardous Materials, 2014, 266: 141-166

[10] Chen Z-m, Wang S-l, Xu J, et al.. Assessment and machine learning prediction of heavy metals fate in mining farmland assisted by Positive Matrix Factorization [J]. Journal of Environmental Management, 2024, 350: 119587

[11] Shi W-c, Li T, Feng Y, et al.. Source apportionment and risk assessment for available occurrence forms of heavy metals in Dongdahe Wetland sediments, southwest of China [J]. Science of the Total Environment, 2022, 815: 152837

[12] Pan Y-x, Chen M, Wang X-t, et al.. Ecological risk, source apportionment, and influencing factors of heavy metals in soil in a typical lead-zinc mining watershed, Guangxi, China [J]. Journal of Environmental Chemical Engineering, 2024, 12(3): 112731

[13] Nascimento S C, Cooke D R, Cracknell M J, et al.. Mineralogical and geochemical characterization of mine tailings in the King river delta, Western Tasmania: Implications for long-term stability of trace elements [J]. Applied Geochemistry, 2025, 184: 106366

[14] Pan J-h, Zhou C-c, Tang M-c, et al.. Study on the modes of occurrence of rare earth elements in coal fly ash by statistics and a sequential chemical extraction procedure [J]. Fuel, 2019, 237: 555-565

[15] Wijaya A R, Khoerunnisa F, Armid A, et al.. The best-modified BCR and Tessier with microwave-assisted methods for leaching of Cu/Zn and their t5⁶⁵Cu/(5⁶⁶Zn for tracing sources in marine sediment fraction [J]. Environmental Technology & Innovation, 2022, 28: 102663

[16] Rauret G, Lopez-Sanchez J F, Sahuquillo A, et al.. Improvement of the BCR three step sequential extraction procedure prior to the certification of new sediment and soil reference materials [J]. Journal of Environmental Monitoring, 1999, 1(157-61

[17] Liu J-q, Zhao J, Du J-p, et al.. Machine learning predicts heavy metal adsorption on iron (oxyhydr) oxides: A combined insight into the adsorption efficiency and binding configuration [J]. Science of the Total Environment, 2024, 950: 175370

[18] Chen H, Li J X, Zhang S W, et al.. Study on the acid - base surface property of themagnetite graphene oxide and its usage for the removal ofradiostrontium from aqueous solution [J]. ract, 2013, 101(12785-794

[19] Li J-j, Cao S, Song W-d, et al.. Visualization and quantification of pore structure in cement tailings waste rock composites using X-ray computed tomography and deep learning [J]. Construction and Building Materials, 2025, 476: 141341

[20] Morales-Pérez A, Moreno-Rodríguez V, Del Rio-Salas R, et al.. Geochemical changes of Mn in contaminated agricultural soils nearby historical mine tailings: Insights from XAS, XRD and, SEP [J]. Chemical Geology, 2021, 573: 120217

[21] Kaur A, Mittal M, Bhatti J S, et al.. A systematic literature review on the significance of deep learning and machine learning in predicting Alzheimer’ s disease [J]. Artificial Intelligence in Medicine, 2024, 154: 102928

[22] Proshad R, Rahim M A, Rahman M, et al.. Utilizing machine learning to evaluate heavy metal pollution in the world’ s largest mangrove forest [J]. Science of the Total Environment, 2024, 951: 175746

[23] Ren X, Mi Z-y, Georgopoulos P G. Comparison of machine learning and land use regression for fine scale spatiotemporal estimation of ambient air pollution: Modeling ozone concentrations across the contiguous United States [J]. Environment International, 2020, 142: 105827

[24] Piatak N M, Seal R RII, Hammarstrom J M. Sequential extraction results and mineralogy of mine waste and stream sediments associated with metal mines in Vermont, Maine, and New Zealand [R], 2007, Reston, Virginia, USA, U.S. Geological Survey

[25] Ma T-f, Luo H-j, Sun J-t, et al.. The effect of heavy precipitation on the leaching of heavy metals from tropical coastal legacy tailings [J]. Waste Management, 2024, 186: 1-10

[26] Yang H-r, Huang K, Zhang K, et al.. Predicting heavy metal adsorption on soil with machine learning and mapping global distribution of soil adsorption capacities [J]. Environmental Science & Technology, 2021, 55(2014316-14328

[27] Verma D, Bhatt U M, Vidyarthi A. A machine learning framework for predictive electron density modelling to enhance 3D NAND flash memory performance [J]. e-Prime-advances in electrical engineering, electronics and energy, 2024, 10: 100790

[28] Barkhordari M S, Zhou N-n, Li K-c, et al.. Interpretable machine learning for predicting heavy metal removal efficiency in electrokinetic soil remediation [J]. Journal of Environmental Chemical Engineering, 2024, 12(6): 114330

[29] Maros M E, Capper D, Jones D T W, et al.. Machine learning workflows to estimate class probabilities for precision cancer diagnostics on DNA methylation microarray data [J]. Nature Protocols, 2020, 15(2479-512

[30] Swift D, Cresswell K, Johnson R, et al.. A review of normalization and differential abundance methods for microbiome counts data [J]. WIREs Computational Statistics, 2023, 15(1e1586

[31] Elkhouly A, Andrew A M, Rahim H A, et al.. Data-driven audiogram classifier using data normalization and multi-stage feature selection [J]. Scientific Reports, 2023, 13(11854

[32] Blokhintsev L D, Savin D A. Study of the influence of different methods of taking into account the coulomb interaction on determining asymptotic normalization coefficients within the framework of exactly solvable model [J]. Physics of Atomic Nuclei, 2021, 84(4401-407 2021

[33] KIM Y S, KIM M K, FU N, et al. Investigating the impact of data normalization methods on predicting electricity consumption in a building using different artificial neural network models [J]. Sustainable Cities and Society, 2024: 105570. DOI: https://doi.org/10.1016/j.scs.2024.105570.

[34] Muhammad A P. Investigating the impact of Min-Max data normalization on the regression performance of k-nearest neighbor with different similarity measurements [J]. Aro-the Scientific Journal of Koya University, 2022, 10: 85-91

[35] Zhang C, Wang J, Li L, et al.. Utilization of earth-to-air heat exchanger to pre-cool/heat ventilation air and its annual energy performance evaluation: A case study [J]. Sustainability, 2020, 12: 8330

[36] Umar M A, Chen Z, Shuaib K, et al.. Effects of feature selection and normalization on network intrusion detection [J]. Data Science and Management, 2025, 8(1): 23-39

[37] GARCÍA J, ENTRENA J, ALESANCO Á. Empirical evaluation of feature selection methods for machine learning based intrusion detection in IoT scenarios [J]. Internet of Things, 2024: 101367. DOI: https://doi.org/10.1016/j.iot.2024.101367.

[38] Periwal N, Arora P, Thakur A, et al.. Antiprotozoal peptide prediction using machine learning with effective feature selection techniques [J]. Heliyon, 2024, 10(16e36163

[39] Xiao H, Liang X, Li H, et al.. Trends in the prevalence of osteoporosis and effects of heavy metal exposure using interpretable machine learning [J]. Ecotoxicology and Environmental Safety, 2024, 286: 117238

[40] Maros M E, Capper D, Jones D T W, et al.. Machine learning workflows to estimate class probabilities for precision cancer diagnostics on DNA methylation microarray data [J]. Nature Protocols, 2020, 15(2479-512

[41] Guyon I, Weston J, Barnhill S, et al.. Gene selection for cancer classification using support vector machines [J]. Machine Learning, 2002, 46(1389-422

[42] Zhou X-z, Wen H-j, Zhang Y-l, et al.. Landslide susceptibility mapping using hybrid random forest with GeoDetector and RFE for factor optimization [J]. Geoscience Frontiers, 2021, 12(5101211

[43] Huang J, Peng Y, Hu L. A multilayer stacking method base on RFE-SHAP feature selection strategy for recognition of driver’s mental load and emotional state [J]. Expert Systems with Applications, 2024, 238: 121729

[44] Chai M-y, He Y-h, Wang J-j, et al.. Predicting creep life of CrMo pressure vessel steel using machine learning models with optimal feature subset selection [J]. International Journal of Pressure Vessels and Piping, 2024, 212: 105349

[45] Lou R-h, Liu W-z, Li R-j, et al.. DeepPhospho accelerates DIA phosphoproteome profiling through in silico library generation [J]. Nature Communications, 2021, 12(16685

[46] Garriga R, Mas J, Abraha S, et al.. Machine learning model to predict mental health crises from electronic health records [J]. Nature Medicine, 2022, 28(61240-1248

[47] Kurniati E O, Zeng H, Latypov M I, et al.. Machine learning for predicting compressive strength of sustainable cement paste incorporating copper mine tailings as supplementary cementitious materials [J]. Case Studies in Construction Materials, 2024, 21: e03373

[48] Malakouti S M, Menhaj M B, Suratgar A A. The usage of 10-fold cross-validation and grid search to enhance ML methods performance in solar farm power generation prediction [J]. Cleaner Engineering and Technology, 2023, 15: 100664

[49] Guo H-n, Liu H-t, Wu S-B. Simulation, prediction and optimization of typical heavy metals immobilization in swine manure composting by using machine learning models and genetic algorithm [J]. Journal of Environmental Management, 2022, 323: 116266

[50] Abba S I, Pham Q B, Saini G, et al.. Implementation of data intelligence models coupled with ensemble machine learning for prediction of water quality index [J]. Environmental Science and Pollution Research International, 2020, 27(3341524-41539

[51] Yao Z-y, Wang Z-c, Wang D-w, et al.. An ensemble CNN-LSTM and GRU adaptive weighting model based improved sparrow search algorithm for predicting runoff using historical meteorological and runoff data as input [J]. Journal of Hydrology, 2023, 625: 129977

[52] Yang D-q, Xiao B-q, Cao M-y, et al.. A new hybrid credit scoring ensemble model with feature enhancement and soft voting weight optimization [J]. Expert Systems with Applications, 2024, 238: 122101

[53] Srinivas A, Mosiganti J P. A brain stroke detection model using soft voting based ensemble machine learning classifier [J]. Measurement: Sensors, 2023, 29: 100871

[54] Choubin B, Moradi E, Golshan M, et al.. An ensemble prediction of flood susceptibility using multivariate discriminant analysis, classification and regression trees, and support vector machines [J]. Science of the Total Environment, 2019, 651: 2087-2096

[55] Dao P B. On Wilcoxon rank sum test for condition monitoring and fault detection of wind turbines [J]. Applied Energy, 2022, 318: 119209

[56] Ma X-d, Guan D-x, Zhang C-s, et al.. Improved mapping of heavy metals in agricultural soils using machine learning augmented with spatial regionalization indices [J]. Journal of Hazardous Materials, 2024, 478: 135407

[57] Bi Z-h, Sun J, Xie Y-t, et al.. Machine learning-driven source identification and ecological risk prediction of heavy metal pollution in cultivated soils [J]. Journal of Hazardous Materials, 2024, 476: 135109

[58] Xie Y-j, Yang F, Liu J, et al.. Prediction of the degradation of organic pollutants by metal-activated peracetic acid using machine learning [J]. Journal of Environmental Chemical Engineering, 2024, 12(5113700

[59] Zhou M, Hu T, Wu M-t, et al.. Rapid estimation of soil Mn content by machine learning and soil spectra in large-scale [J]. Ecological Informatics, 2024, 81: 102615

[60] Yang J-f, Chen Z-g, Wang X-j, et al.. Elucidating nitrogen removal performance and response mechanisms of anammox under heavy metal stress using big data analysis and machine learning [J]. Bioresource Technology, 2023, 382: 129143

[61] Yuan X-z, Suvarna M, Low S, et al.. Applied machine learning for prediction of CO₂ adsorption on biomass waste-derived porous carbons [J]. Environmental Science & Technology, 2021, 55(1711925-11936

[62] Guo L-s, Xu X, Niu C-c, et al.. Machine learning-based prediction and experimental validation of heavy metal adsorption capacity of bentonite [J]. The Science of the Total Environment, 2024, 926: 171986

[63] Guo L-s, Xu X, Wang Q, et al.. Machine learning-based prediction of heavy metal immobilization rate in the solidification/stabilization of municipal solid waste incineration fly ash (MSWIFA) by geopolymers [J]. Journal of Hazardous Materials, 2024, 467: 133682

[64] Zuo R-l, Zhang Z-h, Liang J-h, et al.. The shifting rule analysis of O₂^∲→Eu³⁺ charge transfer band in Re₂Zr₂O₇ via the perspective of electronegativity competition [J]. Journal of Alloys and Compounds, 2025, 1014: 178834

[65] Sanderson R T. Electronegativity and bond energy [J]. Journal of the American Chemical Society, 1983, 105(82259-2261

[66] Zhao J-f, Guo Z-z, Dong J-h, et al.. Exploring the role of pyrite in promoting as attenuation in constructed wetland [J]. Chemical Engineering Journal, 2024, 502: 158055

[67] Tu Z-h, Liang Y-j, Zhou S, et al.. Application of pyrite to water pollutant removal: A review [J]. Journal of Water Process Engineering, 2024, 68: 106375

[68] Wang J-W. A novel ensemble machine learning exposure model system for ground-level ozone at the national scale: A case of mainland China from 2013 to 2020 [J]. Environmental Impact Assessment Review, 2024, 109: 107630

RIGHTS & PERMISSIONS

Central South University