Retention Indices Estimation in Chromatography

Motivated by the requirement of analysis and separation of polycyclic aromatic hydrocarbons, several authors have determined the retention indices of some of these compounds in chromatographic columns. Since the retention indices of many polycyclic aromatic hydrocarbons have not been determined yet, it is desirable to find the mathematical model about the relationships between the value of retention indices and molecular structure of polycyclic aromatic hydrocarbons. Some of these relationships have been studied by PLS. But it can be shown that support vector machine can give mathematical model with better prediction ability. Table 13.3 illustrates the experimental values of 33 polycyclic aromatic hydrocarbons and their molecular parameters [52 5]. Support vector regression has been used for this modeling work. 

The data listed in Table 13.3 are used for model-building by following steps at first KL transformation is carried out, and then SVR-based algorithm is used to make feature selection, then the data file after feature selection is used for mathematical modeling. The same data file is also treated by PLS method. The errors of prediction in LOO cross-validation test are listed in Table 13.4. 

It can be seen that the best result is obtained by SVR computation of the data after KL transformation and feature selection. 

Similar calculation has been also carried out for 32 samples of alkyl-substituted polycyclic aromatic hydrocarbons. Table 13.5 lists the values of errors obtained by different methods. It can be seen that the result of SVR with a data file obtained by SVR-based feature selection is also the best one. 

SVR for feature selection after KL transformation 0.0995 SVR without feature selection 0.112 

---