Solving the spectroscopy interference effects of beta-carotene and lycopene by neural networks

In this study a new computerized approach and linear models (LMs) to solve the UV/vis spectroscopy interference effects of beta-carotene with lycopene analysis by neural networks (NNs) are considered. The data collected (absorbance values) obtained by UV/vis spectrophotometry were transferred into an NN-trained computer for modeling and prediction of output. Such an integrated NN/UV/vis spectroscopy approach is capable of estimating beta-carotene and lycopene concentrations with a mean prediction error 50 times lower than that calculated by the LM/UV/vis spectroscopy approach (without any previous physicochemical knowledge of the process to be modeled).