论文部分内容阅读
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 Artificial Neural Network (ANN) is used to understand the nonlinear dynamic behavior of ore-forming systems and to predict the mineral resources.The Radial Basis Function Neural Network (RBFNN) Which is suitable for the prediction of mineral resources which need a large amount of data synthesis.The prediction of gold mineralization in the southeastern Yunnan area of China with complicated geological conditions is carried out by RBFNN method.The results show that this model can And quickly obtain the information of metallogenic potential.According to the accuracy verification of Re-ceiver Operating Characteristic (ROC) curve, it shows that this model has superior predictive ability.