【摘要】 基于自组织特征映射网络的聚类分析广泛应用于各种聚类分析领域,尝试将此网络应用于干旱特征的聚类分析,使用软件MATLAB7.1对所设计的网络进行学习和训练,经过6000次训练,最终实现了干旱类型的聚类,误差也达到精度要求,所建模型经测试,可以用于分类的预测。SOM网络可以很好地反映和提取气象样本中复杂的信息,分类效果较好,可以在干旱分类中广泛应用。更多还原

【Abstract】 Clustering analysis based on self-organizing feature maps(SOM) network has been widely applied in many areas.In this paper,an attempt was made to apply the network to drought characteristics.The software MATLAB 7.1 was used to train the designed network by 6000 times.The ultimate aim was to the C1uster Analysis of drought types.The model could be used to predict classification by test.The SOM network could reflect the complicated information among each Meteorological stylebooks.The effect of classification was good,and it could be applied on drought classification.更多还原