用神经网络学习识别化探及物探数据  

Learned Classification of Geochemical-Geophysical Data Using a Parallel Massively Neural Network

作　　者：刘瑞林[1] 

机构地区：[1]江汉石油学院物探系

出　　处：《地质与勘探》1992年第2期47-50,共4页Geology and Exploration

摘　　要：用层状神经网络学习识别法对湘南铅、锡矿的化探、物探数据进行了分类研究。结果表明,对铅、锡矿已知样本集的再认正确率均达100%。对待查样本的识别也取得了好效果。A parallel massively neural network was used to classify the geochemical-geophysical data froma Pb-Sn Mine in southern Hunan.Results obtained show that for known samples either of Pb and Snores,the rate of accurate judgement in either case is up to 100%,while for unknown samples it isbetter than that obtained by conventional statistical pattern recognition technique.The parallel massi-vely neural network is effective in learned classification of geophysical-geochemical data and its effe-ctiveness depends only upon the choice of training sets.

关 键 词：神经网络 识别 地化勘探 地球物理勘探 数据 

分 类 号：P62[天文地球—地质矿产勘探]