Artificial neural network modeling of the antioxidant activity of lettuce submitted to different postharvest conditions

Abstract

© 2019 Wiley Periodicals, Inc. Using novel ebb and flow hydroponic system lettuce (Lactuca sativa L.) samples were grown and then stored under different postharvest conditions (greenhouse with a cube, refrigerator with a cube, and refrigerator without a cube). The effect of the postharvest conditions on bioactive compounds profile and antioxidant activity was studied. All samples were subjected to the bioactive compounds composition analysis as well as to the antioxidant activity determination in order to reveal how the postharvest conditions affect them. The experimental data were used for the antioxidant activity modeling by artificial neural network (ANN) approach. The antioxidant activity was determined by DPPH assay and expressed as IC 50 . According to the standard statistical parameters, comparison of the experimentally observed and predicted data, and residuals analysis, generated ANNs could be used for the prediction of the antioxidant activity. Practical applications: Lettuce (Lactuca sativa L.) samples were grown using novel Ebb and Flow hydroponic system. Different postharvest conditions were used for samples storage. Generated models for the prediction of antioxidant activity were obtained using artificial neural network (ANN) modeling.