Comparison of calibration methods with and without feature selection for the analysis of HPLC data

A comparison of two multivariate calibration methods, partial least squares (PLS) and principal component regression (PCR), applied to high-performanc e liquid chromatography (HPLC) data, is presented for the resolution of a p esticide mixture. The data set showed both coeluted peaks and overlapped ab sorption spectra. Besides, there is an additional overlapping between the s ignal of the mobile phase and that of some pesticide. Multivariate calibrat ion models were evaluated using different criteria to choose the optimum nu mber of latent variables. It is shown that PLS yields the best predictive m odels. Furthermore, two methods for selecting regions were applied with the goal to achieve an improved prediction ability in the present multicompone nt determination by HPLC-DAD (diode array detector) with PLS. The selection of regions associated with a large correlation to the concentration and wi th large values in loading-weighs (from PLS) were considered. It is conclud ed that feature selection can also improve the multivariate calibration res ults using chromatographic data.