论文部分内容阅读
应用支持向量回归(Support vector regression,SVR)方法,结合粒子群参数寻优(Particle swarm optimization,PSO)技术,对分子结构参数与分子性能之间的关系进行研究来预测阴离子表面活性剂的临界胶束浓度。并与基于人工神经网络模型的预测结果进行了比较.结果表明:对于相同的训练样本和检验样本,支持向量回归比BPNN模型有更高的预测精度。
The relationship between molecular structure parameters and molecular properties was studied by using Support Vector Regression (SVR) and Particle Swarm Optimization (PSO) to predict the critical plasticity of anionic surfactants Bundle concentration. And compared with the prediction results based on the artificial neural network model.The results show that the support vector regression has higher prediction accuracy than the BPNN model for the same training samples and test samples.