Predicting Shelf Life of MA Packed Paneer Using Feedforward ANN Model

Artificial neural network model was developed for predicting shelf life of MA packed paneer. Levenberg-Marquardt algorithm along with Feedforward backpropagation was used for experimentation. Moisture, titratable acidity, free fatty acids and tyrosine were used as input parameters, while overall acceptability was taken as the output parameter for developing feedforward ANN models. Data was divided into two disjoint sets, viz. 80% used for training and 20% for testing. The combination of 4à45à1 gave the best fitting, suggesting that developed model has the potential for predicting shelf life of MA packed paneer with accuracy.