论文部分内容阅读
为了探讨表面活性剂的增溶性能,计算了表征分子组成和拓扑等特征的148个分子描述符,经属性筛选得到13个描述符,采用支持向量机方法建立了表面活性剂增溶性能分类预测模型。结果表明,该模型预测能力及稳定性良好,5折交叉验证准确率为92.1%,测试集验证准确率为95.1%。用中药皂苷化合物对该模型进行验证,模型验证准确率达93.8%,表明该模型具有良好的推广能力,可为中药增溶性能研究提供指导。
In order to investigate the solubilization performance of surfactants, 148 molecular descriptors were selected to characterize molecular composition and topological characteristics. Thirteen descriptors were screened by attribute selection. The classification of surfactant solubilization performance was predicted by using support vector machine model. The results show that the model has good predictive ability and stability. The accuracy of the 5-fold cross-validation is 92.1% and that of the test set is 95.1%. The model was verified by the saponin compound of Chinese traditional medicine. The accuracy of the model validation was 93.8%, indicating that the model has good promotion ability and can provide guidance for the study of solubilization properties of traditional Chinese medicine.