TPE-SVM Model and SHAP Analysis to Identify Pb-Zn Deposit Types Based on Sphalerite Trace Elements
Chen Zhongyuan
Ren Tao
Zhao Dong
Abstract：This study demonstrates the efficacy of machine learning algorithms in classifying genetic types of Pb-Zn deposits through trace elements in sphalerite.It compiled a comprehensive trace element dataset comprising 3 117 sphalerite samples from 109 globally representative Pb-Zn deposits including MVT,VMS,SEDEX,skarn,and epithermal deposits.Twelve trace elements(Mn,Fe,Co,Cu,Ga,Ge,Ag,Cd,In,Sn,Sb,Pb)were systematically analyzed to develop a Tree-structured Parzen Estimator(TPE)-optimized Support Vector Machine(SVM)classification model.The model demonstrated exceptional discriminative performance on test datasets,achieving accuracy,recall,and F1-score values exceeding 0.97.SHAP(SHapley Additive exPlanations)interpretability analysis revealed Mn,Ge,and Co as critical discriminators among deposit types,providing quantitative insights into elemental controls on genetic classification.The discriminant index system of trace elements in sphalerite established in this paper not only provides a new technical means for the identification of ore genesis,but also provides innovative solutions for complex geological problems such as the analysis of composite metallogenic system and the prediction of concealed ore bodies.
Keywords：sphaleritetrace elementsmachine learningTPE optimization algorithmSHAP algorithmore deposits
Publication Date：2025-11-30
Online Publishing Date：2025-12-15(First online date of this platform, not the publication date of the document)
Pages：15( 4355-4369 )