基于电性拓扑状态指数的烷烃密度定量结构-性质关系研究

QSPR Study on the Density of Alkanes by Using Electrotopological State Index

研究了烷烃密度的定量结构-性质关系。以电性拓扑状态指数为结构描述符,分别用多元线性回归(MLR)和人工神经网络(ANN)建立了结构描述符和密度之间的校正模型。用留一交叉验证和外部测试集验证评价所建立MLR和ANN模型的预测能力。对于MLR模型,这两种验证的均方根相对误差分别为3.37和1.92。对于ANN模型,这两种验证的均方根相对误差为1.06和1.34。这说明建立的MLR和ANN模型都可用于预测烷烃的密度,但ANN模型优于MLR模型。

The quantitative structure property relationship （QSPR） for the density of alkanes was investigated. Electrotopological state （ETS） index was used as the structural descriptor. The relationship between the ETS index and density was modeled with multivariate linear regression （MLR） and artificial neural network （ANN） respectively. Leave one out cross validation （Loo-CV） and external test set validation （EV） were conducted to assess the prediction performance of the developed models. For the MLR model, the root mean square relative error （RMSRE） of Loo-CV and EV is 3.37 and 1.92 respectively. For the ANN model, the RMSRE of Loo-CV and EV is 1.06 and 1,34 respectively. It is demonstrated that there is a quantitative relationship between the ETS index and density. Both MLR and ANN are practicable for modeling this relationship.