| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.22 MB | Adobe PDF |
Advisor(s)
Abstract(s)
Extraction constitutes a vital procedure when attaining bioactive compounds from plant matrices. Conventional extraction using solvents is highly dependent on variables such as time, temperature, solid to liquid ratio or type of solvent, among others, leading to the need for optimising process variables in order to increase the yield.1,2
This research study focuses on the evaluation of total phenolic compounds (TPC) and antioxidant activity (AOA) of strawberry fruits according to different experimental extraction conditions by application of Artificial Neural Networks (ANNs) technique. The experimental data was applied to train ANNs using feed and cascade forward back propagating models by Levenberg-Marquardt and Baysian regulation algorithms. Three independent variables (solvent concentration, volume/mass ratio and extraction time) were used as ANNs inputs whereas the three variables of total phenolic compounds, DPPH and ABTS Antioxidant Activities were considered as ANNs outputs. The results demonstrated that the best neural network cascade and feed forward back-propagation topologies for the prediction of total phenolic compounds and DPPH and ABTS antioxidant activity factors were the 3-9-1, 3-4-4-1 and 3-13-10-1 structures with the training algorithm of trainlm, trainbr, trainlm and threshold functions of tansig-purelin, logsig-tansig-tansig and tansig-tansig-purelin, respectively. The best R2 value for the predication of total phenolic compounds and DPPH and ABTS antioxidant activity factors were 0.9806 (MSE=0.0047), 0.9651(MSE=0.0035) and 0.9756 (MSE=0.00286), respectively. According to the comparison of ANNs, the results showed that cascade forward back propagation network had better performance than feed forward back propagation network for the prediction of TPC as feed forward back propagation network in predicting the DPPH and ABTS antioxidant activity factors had more precision than cascade forward back propagation network. According to the obtained results, it was possible to predict TPC and AOA as a function of extraction time, volume/mass ratio, solvent concentration and volume.
Description
Keywords
Citation
Golpour I, Gonçalves F, Correia P, Guiné R. (2021) Phenolic Compounds and Antioxidant Activity Modeling in Strawberry by using Artificial Neural Networks (ANNs) Technique, Livro de Resumos do XV Encontro de Química dos Alimentos: Estratégias para a Excelência, Autenticidade, Segurança e Sustentabilidade Alimentar, Madeira, Funchal, pp. 107.