摘要
自组织特征映射神经网络(SOM)以无监督方式进行网络训练,具有自组织功能。网络通过自身训练,自动对输入模式进行分类。中药药用价值与其所含微量元素有直接的关系,药材分类是中药质量控制的重要方法。将金银花中微量元素含量作为网络输入,利用自组织特征映射神经网络对不同产地金银花进行分类。结果表明分类效果较好,符合生产实际。
Self-organizing feature map neural network（SOM） trains the network in a way of unsupervised learning and it has the function of self-organising. The network can automatively sort the input pattern by self-training. Traiditional Chinese medicinal value is directly related to trace elements and classification is an important method for quality control. The trace element contents were used as the inputs of network, so the flos lonicerae of different producing area was classified by self-organizing feature map neural network. The results show that the effect of classification is good and according with practical production.
出处
《化学分析计量》
CAS
2013年第2期35-37,共3页
Chemical Analysis And Meterage
基金
四川省中医药管理局资助项目(2008–12)
关键词
自组织特征映射神经网络
金银花
分类
self-organizing feature map neural network
flos lonicerae
classification
