河北地质大学学报2025，Vol.48Issue(4)：20-27,36,9.DOI:10.13937/j.cnki.hbdzdxxb.2025.04.003
基于随机森林算法的南岭东段离子吸附型稀土矿成矿预测
Metallogenic Prediction of Ion-adsorption Rare Earth Deposits in the Eastern Part of Nanling Based on Random Forest Algorithm
摘要
Abstract
Ion-adsorption rare earth ores are a critical source of strategic metals in South China.This paper focuses on the ion-adsorption rare earth deposits in the eastern part of Nanling region.Guided by the endogenic-exogenic metallogenic theory,a multi-factor collaborative metallogenic prediction model is constructed based on ArcGIS platform and random forest algorithm.Nine key ore-controlling variables were extracted by the structural buffer analysis,the normalized difference vegetation index(NDVI),calculation of topographic and geomorphic parameters,including elevation,slope,surface cutting depth,elevation coefficient of variation and topographic relief,and delineation of the distribution range of ore-forming parent rocks.The study area was divided using a 1 000 m×1 000 m grid unit,generating a balanced data set containing 221 positive samples(mine point units)and an equal number of negative samples.The prediction accuracy of the model test set was 77.6%,and the AUC value was 0.94,showing strong generalization ability.The prediction results indicate significant spatial differentiation in metallogenic probability across the study area.Considering the geological background and current exploration status,a total of five metallogenic prospective areas were delineated,classified into A and B levels,including three A-level prospective areas and and two B-level prospecting areas.关键词
离子吸附型稀土矿/随机森林算法/成矿预测Key words
ion-adsorption rare earth type deposit/random forest/metallogenic prediction分类
天文与地球科学引用本文复制引用
李若丹,刘新星,张娟,关子琼,王凯迪,刘美丽,王丽杰..基于随机森林算法的南岭东段离子吸附型稀土矿成矿预测[J].河北地质大学学报,2025,48(4):20-27,36,9.基金项目
自然资源部离子型稀土资源与环境重点实验室基金(2022IRERE105) （2022IRERE105）
国家自然科学基金(41702352) （41702352）
河北地质大学河北省战略性关键矿产协同创新中心开放项目(2024403084) （2024403084）
