modelling some physical characteristics of pomegranate (punica granatum l.) fruit during ripening using artificial neural network

pomegranate is an important iranian-native fruit, with many varieties cultivated. although the volume of data on the importance of pomegranates in human nutrition has increased tremendously in the last years, the physical properties of the pomegranate fruit during fruit maturity have not yet been studied in detail. thus, the present study aimed to evaluate changes in physical characteristics of six pomegranate fruits in three different stages from fruit set to ripening. physical characteristics of pomegranate fruit including length to diameter ratio of fruit and calyx, peel and aril percentage, juice weight and percentage in a whole fruit in ‘aghaye’ (a), ‘faroogh’ (f), ‘rabbab-e-fars’ (rf), ‘shahvare’ (s), ‘shirin-e-bihaste’ (sb) and ‘shirin-e-mohali’ (sm) were investigated. different topologies of the artificial neural network were examined. among different structures, a multilayer feed forward neural network based on 15 neurons in the single hidden layer with transfer function of tangent hyperbolic both in hidden layer and output layer and levenberg-marquardt learning rule was found to be the best model for predicting the physical characteristics of pomegranate fruit from the different cultivars. results indicated that artificial neural network provides a prediction method with high accuracy. the correlation coefficients in the prediction of these physical characteristics were higher than 0.89.