Quantitative analysis of multiple components based on support vector machine (SVM)

In this paper, support vector machine (SVM) is applied to the spectral detection of multiple components to establish a suitable prediction model in the water environment system. Firstly, the mixed fluorescence spectra of four common drugs with extremely similar fluorescence peaks are measured in water quality monitoring. Then, cross-validation is used to select the best parameters of SVM, which are given to obtain the relationship between the fluorescence spectra and corresponding concentration. The results show that the error between the real result and the predicted value is within the allowable error. With the features of effectiveness and robustness, the proposed method could be applied to the quantitative analysis of multiple components.