Abstract
Prospectivity models that quantify spatial associations between predictor variables and mineralization are critical in data-driven mineral prospectivity mapping (MPM). The current prospectivity modelling approaches, however, are limited in terms of meeting both nonlinearity and interpretability, which are two essential factors in evaluating the performance of a model. Along these lines, this work proposed a novel Bayesian decomposition modelling (BDM) method with nonlinear ability and without loss of interpretability. Following the “decomposition-integration” strategy, the BDM method first decomposed the predictor variables by dimension and then transformed them through nonlinear mapping functions (NMFs). All nonlinear mappings of the predictor variables were then linearly integrated to generate the prospectivity model. Throughout the process, the interpretability of the model was maintained via the decomposition of the predictor variables, and the introduction of NMFs ensured better predictive performance compared to methods that only use linear transformations, such as logistic regression and weights of evidence. The BDM method was applied to the Dayingezhuang deposit, a structurally controlled hydrothermal gold deposit in the Jiaodong Peninsula. The results demonstrate that the area under the receiver operating characteristic curve for the classification model was 0.97, and the coefficient of determination of the gold grade predicted by the regression model was 0.91, clearly indicating that the BDM method possesses excellent predictive ability compared to that of several well-known approaches. Moreover, the classification and regression coefficients of the BDM reveal favourable conditions for gold mineralization in Dayingezhuang, including the negative distance (footwall of fault), gentle dip angle, negative divergence of fluid (mineral fluid convergence) and positive volumetric strains (dilation space). Therefore, the proposed BDM method, which combines robust predictive capability and interpretability, is an effective new method for MPM.
Similar content being viewed by others
References
Abedi M, Norouzi G (2012) Integration of various geophysical data with geological and geochemical data to determine additional drilling for copper exploration. J Appl Geophys 83:34–45. https://doi.org/10.1016/j.jappgeo.2012.05.003
Abedi M, Norouzi G, Bahroudi A (2012) Support vector machine for multi-classification of mineral prospectivity areas. Comput Geosci 46:272–283. https://doi.org/10.1016/j.cageo.2011.12.014
Agterberg F, Bonham-Carter G (1999) Logistic regression and weights of evidence modeling in mineral exploration. In: Proceedings of the 28th international symposium on applications of computer in the mineral industry (APCOM), Golden, Colorado, pp 483–490
Agterberg F, Bonham-Carter G (2005) Measuring the performance of mineral-potential maps. Nat Resour Res 14(1):1–17. https://doi.org/10.1007/s11053-005-4674-0
Agterberg F, Bonham-Carter G, Cheng Q, Wright D (1993) Weights of evidence modeling and weighted logistic regression for mineral potential mapping. Computers in Geology-25 years of progress, pp 13–32
Aria M, Cuccurullo C, Gnasso A (2021) A comparison among interpretative proposals for Random Forests. Mach Learn Appl 6:100094. https://doi.org/10.1016/j.mlwa.2021.100094
Barreno M, Cardenas A, Tygar J (2007) Optimal ROC curve for a combination of classifiers. Adv Neural Inf Process Syst 20:57–64
Bonham-Carter G, Agterberg, F, Wright D (1989) Weights of evidence modelling: a new approach to mapping mineral potential. In: Agterberg FP, Bonham-Carter GF (eds) Statistical applications in the earth sciences. Geological Survey of Canada, paper 89-9, pp 171–183
Breiman L (2001) Random forests. Mach Learn 45(1):5–32. https://doi.org/10.1023/A:1010933404324
Bromiley P (2003) Products and convolutions of Gaussian probability density functions density functions. Division of Informatics, Imaging and Data Sciences, School of Health Sciences, University of Manchester, Manchester, M13, 9PT, UK.
Brown W, Gedeon T, Groves D, Barnes R (2000) Artificial neural networks: a new method for mineral prospectivity mapping. Australian J Earth Sci 47(4):757–770. https://doi.org/10.1046/j.1440-0952.2000.00807.x
Carranza EJM (2004) Weights of evidence modeling of mineral potential: a case study using small number of prospects, Abra, Philippines. Nat Resour Res 13(3):173–187. https://doi.org/10.1023/B:NARR.0000046919.87758.f5
Carranza EJM (2011a) From predictive mapping of mineral prospectivity to quantitative estimation of number of undiscovered prospects. Resour Geol 61(1):30–51. https://doi.org/10.1111/j.1751-3928.2010.00146.x
Carranza EJM (2011b) Geocomputation of mineral exploration targets. Comput Geosci 37(12):1907–1916. https://doi.org/10.1016/j.cageo.2011.11.009
Carranza EJM (2017) Publications on geochemical anomaly and mineral potential mapping, and introduction to the special issue of papers in these fields. Nat Resour Res 26(4):379–410. https://doi.org/10.1007/s11053-017-9348-1
Carranza EJM, Hale M (2002) Wildcat mapping of gold potential, Baguio District, Philippines. Trans Inst Min Metall 111:B100–B105
Chen J, Mao X, Liu Z, Deng H (2020) Three-dimensional metallogenic prediction based on random forest classification algorithm for the Dayingezhuang Gold Deposit. Geotecton Metallog 44(2):231–241. https://doi.org/10.16539/j.ddgzyckx.2020.02.007
Cheng Q, Agterberg F (1999) Fuzzy weights of evidence and its application in mineral potential mapping. Nat Resour Res 8:27–35
Cortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20(3):273–297. https://doi.org/10.1023/A:1022627411411
Deng J, Yang L, Sun Z, Wang J, Wang Q, Xin H, Li X (2003) A metallogenic model of gold deposits of the Jiaodong granite-greenstone belt. Acta Geol Sin Engl Ed 77(4):537–546
Deng J, Wang Q, Wang L, Liu H, Yang L, Zhang J (2011) A multifractal analysis of mineralization characteristics of the Dayingezhuang disseminated-veinlet gold deposit in the Jiaodong gold province of China. Ore Geol Rev 40(1):54–64. https://doi.org/10.1016/j.oregeorev.2011.05.001
Deng J, Yang L, Li R, Groves D, Santosh M, Wang Z, Sai S, Wang S (2019) Regional structural control on the distribution of world-class gold deposits: an overview from the Giant Jiaodong Gold Province, China. Geol J 54(1):378–391. https://doi.org/10.1002/gj.3186
Deng J, Yang L, Groves D, Zhang L, Qiu K, Wang Q (2020) An integrated mineral system model for the gold deposits of the giant Jiaodong province, eastern China. Earth-Sci Rev 208:1032. https://doi.org/10.1016/j.earscirev.2020.103274
Deng H, Huang X, Mao X, Yu S, Chen J, Liu Z, Zou Y (2022a) Generalized mathematical morphological method for 3D shape analysis of geological boundaries: application in identifying mineralization-associated shape features. Nat Resour Res 31(4):2103–2127. https://doi.org/10.1007/s11053-021-09975-6
Deng H, Zheng Y, Chen J, Yu S, Xiao K, Mao X (2022b) Learning 3D mineral prospectivity from 3D geological models using convolutional neural networks: application to a structure-controlled hydrothermal gold deposit. Comput Geosci 161:105074. https://doi.org/10.1016/j.cageo.2022.105074
Freund Y, Schapire R (1997) A decision-theoretic generalization of on-line learning and an application to boosting. J Comput Syst Sci 55(1):119–139
Ghezelbash R, Maghsoudi A, Carranza EJM (2020) Sensitivity analysis of prospectivity modeling to evidence maps: Enhancing success of targeting for epithermal gold, Takab district, NW Iran. Ore Geol Rev 120:103394. https://doi.org/10.1016/j.oregeorev.2020.103394
Goldfarb R, Groves D (2015) Orogenic gold: common or evolving fluid and metal sources through time. Lithos 233:2–26. https://doi.org/10.1016/j.lithos.2015.07.011
Goldfarb RJ, Santosh M (2014) The dilemma of the Jiaodong gold deposits: are they unique? Geosci Front 5(2):139–153. https://doi.org/10.1016/j.gsf.2013.11.001
Goodfellow I, Bengio Y, Courville A (2016) Deep learning. MIT Press, Cambridge
Guidotti R, Monreale A, Ruggieri S, Turini F, Giannotti F, Pedreschi D (2018) A survey of methods for explaining black box models. ACM Comput Surv (CSUR) 51(5):1–42. https://doi.org/10.1145/3236009
Huang J, Smola A, Gretton A, Borgwardt K, Schölkopf B (2006) Correcting sample selection bias by unlabeled data. In: Advances in neural information processing systems: proceedings of the 2004 conference
Journel A, Huijbregts C (1978) Mining geostatistics. Academic Press, London
Konstantinov A, Utkin L (2021) Interpretable machine learning with an ensemble of gradient boosting machines. Knowl-Based Syst 222:106993. https://doi.org/10.1016/j.knosys.2021.106993
Li X, Yuan F, Zhang M, Jia C, Jowitt S, Ord A, Zheng T, Hu X, Li Y (2015) Three-dimensional mineral prospectivity modeling for targeting of concealed mineralization within the Zhonggu iron orefield, Ningwu Basin, China. Ore Geol Rev 71:633–654. https://doi.org/10.1016/j.oregeorev.2015.06.001
Li H, Geng K, Zhuo C, Liang T (2016) Tectonic setting and mineralization of the Jiaodong Gold Deposit. Geological Publishing House, Beijing, pp 85–244 (in Chinese)
Li S, Chen J, Liu C, Wang Y (2021a) Mineral Prospectivity prediction via convolutional neural networks based on geological Big Data. J Earth Sci 32(2):327–347. https://doi.org/10.1007/s12583-020-1365-z
Li T, Zuo R, Xiong Y, Peng Y (2021b) Random-drop data augmentation of deep convolutional neural network for mineral prospectivity mapping. Nat Resour Res 30(1):27–38. https://doi.org/10.1007/s11053-020-09742-z
Li T, Zuo R, Zhao X, Zhao K (2022) Random-drop data augmentation of deep convolutional neural network for mineral prospectivity mapping. Ore Geol Rev 142:104693. https://doi.org/10.1016/j.oregeorev.2022.104693
Liu L, Zhang Y (2007) Numerical modeling of the coupled mechanical and hydrological processes during deformation and mineralization in the Mount Isa Block, Australia. Resour Geol 57(3):283–300. https://doi.org/10.1111/j.1751-3928.2007.00023.x
Liu L, Zhao Y, Zhao C (2010) Coupled geodynamics in the formation of Cu skarn deposits in the Tongling-Anqing district, China: computational modeling and implications for exploration. J Geochem Explor 106(1–3):146–155. https://doi.org/10.1016/j.gexplo.2010.01.002
Liu Z, Chen J, Mao X, Tang L, Yu S, Deng H, Wang J, Liu Y, Li S, Bayless, (2021a) Spatial association between orogenic gold mineralization and structures revealed by 3D prospectivity modeling: a case study of the Xiadian gold deposit, Jiaodong Peninsula, China. Nat Resour Res 30(6):3987–4007. https://doi.org/10.1007/s11053-021-09956-9
Liu Z, Hollings P, Mao X, Lawley C, Yang B, Tang L (2021b) Metal remobilization from country rocks into the Jiaodong-type orogenic gold systems, Eastern China: New constraints from scheelite and galena isotope results at the Xiadian and Majiayao gold deposits. Ore Geol Rev 134:104126. https://doi.org/10.1016/j.oregeorev.2021.104126
Liu Z, Mao X, Jedemann A, Bayless R, Deng H, Chen J, Xiao K (2021c) Evolution of pyrite compositions at the Sizhuang gold deposit, Jiaodong Peninsula, Eastern China: implications for the genesis of Jiaodong-type orogenic gold mineralization. Minerals 11:344. https://doi.org/10.3390/min11040344
Liu Z, Mao X, Wang F, Tang L, Chen G, Chen J, Deng H (2021d) Deciphering the anomalous Ag enrichment recorded by galena in the Dayingezhuang Au(-Ag) deposit, Jiaodong Peninsula, Eastern China. Trans Nonferr Met Soc China. https://doi.org/10.1016/S1003-6326(21)65768-0
Lou Y, Caruana R, Gehrke J (2012) Intelligible models for classification and regression. In: Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining, pp 150–158
Macedo I, Gois J, Velho L (2011) Hermite radial basis functions implicits. Comput Graph Forum 30(1):27–42. https://doi.org/10.1111/j.1467-8659.2010.01785.x
Mao X, Ren J, Liu Z, Chen J, Tang L, Deng H, Bayless R, Yang B, Wang M, Liu C (2019) Three-dimensional prospectivity modeling of the Jiaojia-type gold deposit, Jiaodong Peninsula, Eastern China: a case study of the Dayingezhuang deposit. J Geochem Explor 203:27–44. https://doi.org/10.1016/j.gexplo.2019.04.002
Palczewska A, Palczewski J, Marchese Robinson R, Neagu D (2014) Interpreting random forest classification models using a feature contribution method. In: Integration of reusable systems, pp 193–218. https://doi.org/10.1007/978-3-319-04717-1_9
Pan S, Yang Q (2010) A survey on transfer learning. IEEE Trans Knowl Data Eng 22(10):1345–1359. https://doi.org/10.1109/TKDE.2009.191
Porwal A, Carranza EJM, Hale M (2003) Artificial neural networks for mineral-potential mapping: a case study from Aravalli Province, Western India. Nat Resour Res 12(3):155–171. https://doi.org/10.1023/A:1025171803637
Porwal A, Carranza EJM, Hale M (2004) A hybrid neuro-fuzzy model for mineral potential mapping. Math Geol 36(7):803–826. https://doi.org/10.1023/B:MATG.0000041180.34176.65
Porwal A, Gonzalez-Alvarez I, Markwitz V, McCuaig T, Mamuse A (2010) Weights-of-evidence and logistic regression modeling of magmatic nickel sulfide prospectivity in the Yilgarn Craton, Western Australia. Ore Geol Rev 38(3):184–196. https://doi.org/10.1016/j.oregeorev.2010.04.002
Raina R, Shen Y, Ng A, McCallum A (2003) Classification with hybrid generative/discriminative models. In: Proceeding of the NIPS
Rodriguez-Galiano V, Sanchez-Castillo M, Chica-Olmo M, Chica-Rivas M (2015) Machine learning predictive models for mineral prospectivity: an evaluation of neural networks, random forest, regression trees and support vector machines. Ore Geol Rev 71:804–818. https://doi.org/10.1016/j.oregeorev.2015.01.001
Sagar D, Cheng Q, Agterberg F (2018) Handbook of mathematical geosciences: fifty years of IAMG. Springer Nature, Berlin, p 914
Song M, Yi P, Xu J, Cui S, Shen K, Jiang H, Yuan W, Wang H (2012) A step metallogenetic model for gold deposits in the northwestern Shandong Peninsula, China. Sci China Earth Sci 55(6):940–948. https://doi.org/10.1007/s11430-012-4366-7
Song M, Li S, Santosh M, Zhao S, Yu S, Yi P, Cui S, Lv G, Xu J, Song Y, Zhou M (2015) Types, characteristics and metallogenesis of gold deposits in the Jiaodong Peninsula, Eastern North China Craton. Ore Geol Rev 65:612–625. https://doi.org/10.1016/j.oregeorev.2014.06.019
Stein M (1999) Interpolation of spatial data: some theory for kriging. Springer, New York
Wang J, Mao X, Peng C, Chen J, Deng H, Liu Z, Wang W, Fu Z, Wang C (2023) Three-dimensional refined modelling of deep structures by using the level set method: application to the Zhaoping detachment fault, Jiaodong Peninsula, China. Math Geosci 55:229–262. https://doi.org/10.1007/s11004-022-10031-z
Xiao F, Wang K, Hou W, Wang Z, Zhou Y (2020) Prospectivity mapping for porphyry Cu-Mo mineralization in the Eastern Tianshan, Xinjiang, Northwestern China. Nat Resour Res 29(1):89–113. https://doi.org/10.1007/s11053-019-09486-5
Xiao F, Chen W, Wang J, Erten O (2022) A hybrid logistic regression: gene expression programming model and its application to mineral prospectivity mapping. Nat Resour Res 31(4):2041–2064. https://doi.org/10.1007/s11053-021-09918-1
Xie S, Mao X, Liu Z, Deng H, Chen J, Xiao K (2022) Determining the paleostress regime during the mineralization period in the Dayingezhuang Orogenic Gold Deposit, Jiaodong Peninsula, Eastern China: insights from 3D numerical modeling. Minerals 12(5):505. https://doi.org/10.3390/min12050505
Xiong Y, Zuo R (2022) Robust feature extraction for geochemical anomaly recognition using a stacked convolutional denoising autoencoder. Math Geosci 54(3):623–644. https://doi.org/10.1007/s11004-021-09935-z
Yang L, Deng J, Wang Z, Zhang L, Guo L, Song M, Zheng X (2014) Mesozoic gold metallogenic system of the Jiaodong gold province, eastern China. Acta Petrol Sin 30(9):2447–2467
Yang L, Deng J, Wang Z, Zhang L, Goldfarb R, Yuan W, Weinberg R, Zhang R (2016) Thermochronologic constraints on evolution of the Linglong Metamorphic Core Complex and implications for gold mineralization: a case study from the Xiadian gold deposit, Jiaodong Peninsula, eastern China. Ore Geol Rev 72:165–178. https://doi.org/10.1016/j.oregeorev.2015.07.006
Yuan F, Li X, Zhang M, Jowitt S, Jia C, Zheng T, Zhou T (2014) Three-dimensional weights of evidence-based prospectivity modeling: a case study of the Baixiangshan mining area, Ningwu Basin, Middle and Lower Yangtze Metallogenic Belt, China. J Geochem Explor 145:82–97. https://doi.org/10.1016/j.gexplo.2014.05.012
Zhao X, Wu Y, Lee D, Cui W (2019) iForest: Interpreting random forests via visual analytics. IEEE Trans Vis Comput Graph 25(1):407–416. https://doi.org/10.1109/TVCG.2018.2864475
Zhou X, Yang J, Zhang L (2003) Metallogenesis of superlarge gold deposits in Jiaodong region and deep processes of subcontinental lithosphere beneath North China Craton in Mesozoic. Sci China Ser D-Earth Sci 46:14–25
Zuo R, Carranza EJM (2011) Support vector machine: a tool for mapping mineral prospectivity. Comput Geosci 37(12):1967–1975. https://doi.org/10.1016/j.cageo.2010.09.014
Zuo R, Xu Y (2022) Graph deep learning model for mapping mineral prospectivity. Math Geosci 55:1–21. https://doi.org/10.1007/s11004-022-10015-z
Zuo R, Kreuzer O, Wang J, Xiong Y, Zhang Z, Wang Z (2021) Uncertainties in GIS-based mineral prospectivity mapping: key types, potential impacts and possible solutions. Nat Resour Res 30(5):3059–3079. https://doi.org/10.1007/s11053-021-09871-z
Acknowledgements
This research was funded by the National Natural Science Foundation of China (Nos. 42030809, 72088101, 41972309, and 42072325).
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
Here, the product of powers of two univariate Gaussian probability density functions (PDFs) is discussed based on Bromiley (2003). Let \(f(x)\) and \(g(x)\) be Gaussian PDFs with arbitrary means \({\mu }_{f}\) and \({\mu }_{g}\) and standard deviations \({\sigma }_{f}\) and \({\sigma }_{g}\), respectively
By supposing that \(a\) and \(b\) are power exponents of \(f(x)\) and \(g(x)\), respectively, their product is
Next, the term in the exponent is examined
Similar to Bromiley (2003), we denote the variance \({{\sigma }_{fg}}^{2}\) and mean \({\mu }_{fg}\) by
By supposing that \(\epsilon \) is the term required to complete the square in \(\beta \)
Adding this term to \(\beta \) gives
Simplifying, the above formulas are
Substituting back into Eq. 41 gives
Multiplying by \({\sigma }_{fg}/{\sigma }_{fg}\) and rearranging gives
Therefore, the product of the powers of two Gaussian PDFs \({f\left(x\right)}^{a}\) and \({g\left(x\right)}^{b}\) is also a scaled Gaussian PDF. The variance\({{\sigma }_{fg}}^{2}\), the mean \({\mu }_{fg}\), and the scaling factor \({S}_{fg}\) can be written more conveniently as
Finally, Eq. 48 was extended to the product of powers of n univariate Gaussian PDFs according to the proof of Bromiley (2003)
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Mao, X., Wang, J., Deng, H. et al. Bayesian Decomposition Modelling: An Interpretable Nonlinear Approach for Mineral Prospectivity Mapping. Math Geosci 55, 897–942 (2023). https://doi.org/10.1007/s11004-023-10067-9
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11004-023-10067-9