矿床品位估值网络泛化能力研究

    GENERALIZATION ABILITY OF NEURAL NETWORK FOR ESTIMATING DEPOSIT GRADE

    • 摘要: 文中提出了用于矿床品位估值神经网络的泛化能力概念,应用矿山的实例对网络结构、样本数量和样本顺序对泛化能力的影响进行了实验研究,提出了改善估值网络泛化能力的措施。

       

      Abstract: This paper presents the generalization ability of neural network for estimating deposit grade.It is proved that the structure of neural network,sample number and sample sequence have the influence on the ge neralization ability by the test.The method to improve the generalization ability are put forward.

       

    /

    返回文章
    返回