论文部分内容阅读
报道了函数连接型网络（ＦＬＮ）用于维生素Ｂ族四组分同时测定。采用相关系数和标准偏差从原始紫外光谱数据中挑选１１ 个波长点供网络处理。在函数连接型网络中，非线性输入模式得到了增强， 并使用了推广的δ学习规则。预测结果极好， 其相关系数和标准偏差分别为０．９９９０４ 和０．２６８８５。
Functional connectivity networks (FLNs) were reported for the simultaneous determination of four components of vitamin B family. The correlation coefficient and standard deviation were used to select 11 wavelength points from the original UV spectrum data for network processing. In functionally connected networks, the nonlinear input mode is enhanced and a generalized δ learning rule is used. The prediction results were excellent with correlation coefficients and standard deviations of 0.99904 and 0.26885, respectively.