SOM网络在黑龙江省西北部金矿预测中的应用  

Application of Self-Organizing Feature Map(SOM) neural network on gold mineral resources prediction in northeast Heilongjiang province

作　　者：王满[1] 薛林福[1] 

机构地区：[1]吉林大学地球科学学院

出　　处：《黄金》2008年第11期12-15,共4页Gold

摘　　要：在深入研究分析黑龙江省西北部金矿成矿地质背景的基础上,以地理信息系统为平台(GIS),建立了地质、物探、化探和遥感数据库,用字符串表示研究区地质点的各种地质信息组合,利用自组织特征映射(SOM)神经网络对其进行模式分类,圈定了金矿成矿有利区,这一研究对该地区今后的矿产资源预测与评价有重要的参考价值。Based on the comprehensive analysis on gold mineralization geological background in northeast Heilongjiang province, geological, geophysical, geochemical and remote sensing databases have been established with GIS platform. Self-Organizing Feature Map （SOM） neural network is used to classify the character strings presenting various geological information in a certain geological point, and then advantaged gold mineralization zone are figured out. The results are of great reference value to mineral resources prediction and evaluation in the area.

关 键 词：SOM网络 GIS 成矿预测 金矿 黑龙江省西北部 

分 类 号：P618.51[天文地球—矿床学] P628[天文地球—地质学]