Quality of probiotic yoghurt Prediction Using Intelligent Computing Expert System Models

Over the last few years, the use of artificial neural networks (ANNs) has increased in many areas of food science and artificial neural network have been applied to many Quality Control problems. The aim of this study is to develop and optimize artificial neural network models for forecast of viability of probiotic bacteria in yogurt. In this study, A Levenberg-Marquardt algorithm trained feedforward artificial neural was used as a model. Several parameters such as titratable acidity, pH, syneresis, color and bacterial counts (Streptococcus thermophilus, L. delbrueckii ssp. bulgaricus Lactobacillus acidophilus and the viability of probiotic bacteria are considered. The efficiency of the developed model was assessed through a calculate the Mean Square Error (MSE) and value of R2. The transfer function of the neural network model is effective in performance. The two tangent sigmoid and linear transfer function model is used to optimize pre-processing andwere compared. According to the results, the tangent sigmoid stimulation with 10 neurons in the hidden layer Here, the highest coefficients of determination on days 1, 7 and 14 during storage were (0.997, 0.994, 0.993) and the mean square error (MSE) 0.00056, 0.00075 and 0.00084 respectively. The overall accuracy of most models suitable for predicting viability of probiotic bacteria on during storage. Therefore, the best structure of a neural network to predict survival was 34-10-3.