Application of artificial neural network to simulate phenolic content and antioxidant activity during spontaneous fermentation of Carica papaya leaf / Nur Syahira Mohammad and Mohamad Sufian So’aib

Artificial Neural Network (ANN) was used to simulate the phenolic content and antioxidant activity during spontaneous fermentation of Carica papaya leaf. The aim of the present research project is to develop ANN models for prediction of phenolic content and antioxidant activity of Carica papaya leaf during spontaneous fermentation and to compare the statistical performance of different ANN architecture for the prediction of phenolic content and antioxidant activity. Data used to derive and validate the model was obtained from the experiment. The input of the ANN model is volume and day of fermentation, while the output is the phenolic content and antioxidant activity. Trial and error method were used to develop the ANN model. The transfer function used in this research project was hyperbolic tangent sigmoid with the Levenberg- Marquadt algorithm training function. The ANN architecture was the multilayer feed-forward structure with backpropagation training algorithm used for computing biases and weights. The performance of ANN model was being evaluated by correlation coefficient (R) and mean square error (MSE). The neural network model with minimum MSE and maximum R value was considered to be the best ANN. The best topology for Antioxidant activity of Carica Papaya leaf is 2-12-12-1 with low MSE value which is 0.0044367. The best topology for Phenolic content of Carica Papaya leaf is 2-11-11-1 with low MSE value which is at 0.00024449.