论文部分内容阅读
本文采用相似性计算方法将化合物结构信息变量转换为相似系数变量，并以相似系数作为人工神经网络输入参数，预测多元取代苯的疏水分配系数、硝化反应产率分布和烷基苯的汽化焓．结果与文献值符合较好．
In this paper, the similarity calculation method was used to convert compound structural information variables into similarity coefficient variables. The similarity coefficient was used as the input parameter of artificial neural network to predict the hydrophobic partition coefficient, the yield distribution of nitration reaction and the enthalpy of vaporization of alkyl benzene. The result is in good agreement with the literature.