Comparison of Multiway Calibration Techniques in Determination of RNA Nucleotides
Page: 993
aDepartment of Analytical Chemistry, Faculty of Science, Comenius University, Bratislava, and bDepartment of Automatic Control Systems, Faculty of Electrical Engineering and Information Technology, Slovak University of Technology
The quality of several multiway calibration models of spectral data of four RNA nucleotides with different noise is compared. An own genetic algorithm for optimization of the number of wavelengths in spectral matrix for multiple linear regression methods (MLR) was used. Useful analytical information was obtained only for data with noise 0 < 0.02. The prediction error of analyte concentration using a combination of MLR and genetic algorithm was decreased by ca. 50 %.