Variables Selection in the Ultraviolet, Visible and Near Infrared Range for Calibration of a Mixture of Vegetable Oils by Absorbance Spectra

The aim of the work was a multivariate calibration concentration unrefined sunflower oil, considered as adulteration, in mixture with flaxseed oil. relevance study is due to need develop simple and effective method for detecting falsification oil which superior content essential polyunsaturated fatty acids olive A few works only are devoted identifying adulteration unlike oil.Multivariate carried out using model based on principal component analysis, cluster analysis projection latent structures absorbance spectra UV, visible near IR ranges. Calibration uses three methods spectral variables selection: successive projections algorithm, searching combination moving window, ranking by correlation coefficient.The application coefficient window makes it possible reduce value root mean square error prediction from 0.63 % wideband 0.46 %, 0.50 0.03 respectively.The developed ranges selection