摘要:把地质大数据和人工智能技术引入矿产资源定量评价及成矿预测体系中，提高了海量地质数据的有效信息挖掘，弥补了传统方法的不足。本文基于白象山矿区基础地质资料和物化探成果资料，利用三维地质体建模技术和三维空间分析技术，量化三维控矿因素，建立了一种基于CART 算法的三维成矿预测模型。通过在白象山矿区的实验表明：该模型能较好的定位已知矿体，并且预测出在已知矿体北部、东部、东北部、西部、南部和东南部具有较高的成矿概率，可圈定找矿靶区。该模型将地质大数据应用于找矿勘探工作，具有纯数据驱动、预测精度高、预测结果可靠等优点。研究发现，该模型的预测效果与训练数据集的数量、矿控因素提取、决策树深度等有关。?
关键词: 大数据/
机器学习/
CART算法/
三维建模/
成矿预测
Abstract:Applying the geology big data and artificial intelligence technology to the quantitative evaluation of mineral resources and metallogenic prediction systems has improved the effective information mining of massive geological data and made up for the deficiencies of traditional methods. Based on the basic geological data and geophysical and geochemical prospecting data of the Baixiangshan mining area, this paper used the three-dimensional geological modeling techniques and three-dimensional spatial analysis techniques to build a 3D geological body model and quantify three-dimensional ore-controlling factors, establishing a three-dimensional mineralization prediction model based on the CART algorithm. The experiments in the Baixiangshan mining area show that the model can better localize known orebodies and predict a higher probability of mineralization in the northern, eastern, northeastern, western, southern and southeastern regions of the known orebodies. Those areas can be used to define the prospecting targets. This model that applying the geology data to prospecting and the exploration work has the advantages of pure data drive, high prediction accuracy, and reliable prediction results. At the same time the study has found that the prediction effect of the model is related to the number of training data sets, extraction of mining control factors, and decision tree depth.??
Key words:Big data/
Machine learning/
CART algorithm/
3D modeling/
Metallogenic prediction
PDF全文下载地址:
http://www.dzkx.org/data/article/export-pdf?id=geology_11494
