摘要
采用人工网络神经法(Artificial Neural Network,ANN)有助于理解成矿系统的非线性动力学行为和对矿产资源进行预测.其中的径向基神经网络(Radial Basis Function Neural Network,RBFNN)具有优秀的逼近特性,优化过程简单,训练速度快,适合于需要大量数据综合的矿产预测.采用RBFNN方法对成矿地质条件复杂的中国滇东南地区开展金矿成矿预测.研究结果表明,该模型能快速获取成矿潜力信息.通过采用受试者工作特征(Re-ceiver Operating Characteristic,ROC)曲线进行精度验证,表明该模型具有优越的预测能力.
The use of artificial neural network（ANN） helps to understand ore-forming system of the nonlinear dynamical behaviors and prediction of mineral resources.RBFNN（Radial Basis Function Neural Network,RBFNN） has excellenl approximation property.Its optimizing process is simple,training process is fast,which suitable for the needs of integrating huge numbers of data together for mineral prospectivity mapping.In this paper,RBFNN model is employed to gold prospectivity mapping in Southeastern Yunnan,China.The experimental results show that the model can quickly obtain the information of patentiality of gold deposits.ROC（receiver operating characteristic） curve shows the accuracy of this model is excellent.
出处
《浙江大学学报（理学版）》
CAS
CSCD
北大核心
2011年第3期354-361,共8页
Journal of Zhejiang University（Science Edition）
关键词
矿产潜力预测
人工神经网络
径向基神经网络
ROC验证
金矿
滇东南
prospectivity mapping
artificial neural network（ANN）
radial basis function neural network（RBFNN）
receiver operating characteristic（ROC） evaluation
gold deposit
southeastern Yunnan
