Utilizing Artificial Neural Networks for Predictive Modeling Physicochemical Attributes in Maltodextrin-Coated Grapes with Potassium Carbonate and Pyracantha Extract in Storage

Artificial neural networks (ANN) can be used as a nondestructive method for estimating the shelf life and quality attributes of fruits and vegetables. In this research, in order to model the storage process of fruit grapes (Vitis vinifera cv. Rishbaba) coated with maltodextrin, including different levels of potassium nanocarbonate (0 and 2%) and pyracantha extract (0 and 1.5%), artificial neural network was used. After applying these coatings, the fruits were stored for 60 days in a cold storage with a temperature of -1°C and a relative humidity of 90%. Weight loss percentage, Titrable acidity (TA), pH, texture firmness, color index (a*) and general acceptance of fruit grapes were investigated. Artificial neural networks were used to predict changes. By examining different networks, the feedforward backpropagation network with 3-10-6 topologies with coefficient of determination (R²) greater than 0.988 and mean square error (MSE) less than 0.005 and by using hyperbolic sigmoid tangent activation function, resilient learning pattern and 1000 learning process were determined as the best neural method. On the other hand, the results of the optimized models showed that this model had the highest and lowest accuracy for predicting the weight loss percentage (R2= 0.9975) and a* (R2= 0.5671) of the samples respectively.