说明：  针对普通神经网络模型确定滑坡易发性评价指标权重及易发性制图精度不高的问题，提出了一种新的 权重确定方法和滑坡易发性评价模型。将BP 神经网络模型和MIV 理论相结合，获取最优隐藏节点数，优化 神经网络模型。在此基础上，综合BP 神经网络连接矩阵和MIV 值确定滑坡易发性评价指标权重，构建滑坡 评价模型。将评价模型应用于龙南县滑坡易发性制图，并利用ＲOC 曲线对评价结果进行了检验。结果表明: MIV － BP 模型具有较高的精度( AUC = 0． 820 4) ，在滑坡空间预测中具有更高的准确性和较大的应用潜力。
(In view of the problem that the common neural network model determines the weight of the evaluation index of landslide susceptibility and the mapping accuracy is not high, a new method is proposed Weight determination method and evaluation model of landslide susceptibility. By combining BP neural network model with MIV theory, the number of hidden nodes is optimized Neural network model. On this basis, the weight of landslide susceptibility evaluation index is determined by combining BP neural network connection matrix and MIV value to construct landslide Evaluation model. The evaluation model is applied to the mapping of landslide susceptibility in Longnan County, and the ROC curve is used to test the evaluation results. The results show that: Miv-bp model has higher accuracy (AUC = 0.8204), higher accuracy and greater application potential in landslide spatial prediction.)