人工神经网络方法预测气相色谱保留指数

Ptediction of Gas Chromatographic Retention Index by Neural Network

用误差反向传播（ＢＰ）的人工神经网络（ＡＮＮ）模型及分子结构描述码作为输入特征参数，预测气相色谱保留指数．研究了链烷烃、环脂烃、烯烃及醇、酯、醚等３００个化合物，预测结果平均相对误差不大于２．８３％．

A back-propagation neural network model has been trained to predict the gaschromatographic retention indices.A group of molecular descriptors which are proposed torepresent molecular branching,cyclization and unsaturation has been used as inputted struc-tural parameters.In 300 acyclic and cyclic alkanes,alkenes,alcohols,esters,ketones andethers studied,the mean relative error of the predicted results was less than 2.83%.