prediction of methyl salicylate effects on pomegranate fruit quality and chilling injuries using adaptive neuro-fuzzy inference system and artificial neural network

Adaptive neuro-fuzzy inference system (anfis) and genetic algorithm–artificial neural network (ga-ann) were investigated for predicting methyl salicylate (mesa) effects on chilling injuries and quality changes of pomegranate fruits during storage. fruits were treated with mesa at three concentrations(0, 0.01 and 0.1 mm) and stored under chilling temperature for 84 days. anfis and ga-ann models were used to predict the effect of mesa application and storage time (0, 14, 28, 42, 56, 70 and 84 days) on chilling injuries, quality parameters and physiological changes of pomegranate during storage. the ga-ann and anfis were fed with 2 inputs of mesa and time. the developed ga–ann, which included 20 hidden neurons, could predict physiological changes and quality parameters of pomegranate fruit (weight loss, ph, titratable acidity, chilling injury index, ion leakage, ethylene, respiration, polyphenols, anthocyanins, total antioxidant activity) with average correlation coefficient of 0.89. the overall agreement between anfis predictions and experimental data was also significant (r=0.87).in addition, sensitivity analysis results showed that storage time was the most sensitive factor for prediction of mesa effects on pomegranate fruit quality attributes during postharvest storage.