Work

Improving freshness prediction in frozen fish burgers: A comparative study of propolis additives using ANN and RSM models

Year: 2025
Type: article
Abstract: The freshness of fish burgers (FBs) declines during frozen storage. Here, we assessed FB freshness using quality control indicators (QCIs), including peroxide value (PV), total volatile basic nitrogen (TVB-N), and total viable count (TVC). Two predictive models were compared, one based on response surface methodology (RSM) and the other on an artificial neural network (ANN). Their accuracy was evaluated using frozen FBs that incorporated different concentrations of freeze-dried propolis (FDP; 0 %–0.4 %) and stored for various durations (0, 30, 60, and 90 days). Both FDP and storage time (ST) had significant effects ( p ​< ​0.01) on the quality control indicators (QCIs) of frozen FBs, with ST having a more significant effect than FDP on the QCI changes. A numerical optimization process determined that the optimal values of ST and FDP were approximately 27 days and 0.30 ​g/[100 ​g of fish paste], respectively. The coefficient of determination (R 2 ) values for the QCIs of frozen FBs in the ANN model were 0.9657 for PV, 0.9753 for TVB-N, and 0.9872 for TVC. These values were slightly lower in the RSM model, 0.9717 for PV, 0.9603 for TVB-N, and 0.9861 for TVC. Overall, the ANN model with a 2-13-3 topology (13 neurons in the first hidden layer) showed greater potential for prediction of FB quality during frozen storage and was found to be the more efficient method. • Two modeling approaches (RSM and ANN) for predicting the freshness of fish burgers were developed. • Quality control indicators (QCIs) are affected by both the concentrations of freeze-dried propolis and the storage period. • Both modeling approaches could reasonably forecast the QCIs of fish burgers during storage.
Language: English
Open Access status: gold
Estimated APC paid: $1,600
OpenAlex ID: W4407849388
FWCI: 4.075
Cites: 32
Cited by: 7
Related to: 10