Spray drying techniques are one of the methods to preserve and extend the shelf-life of coconut milk. The objective of this research was to create a particle swarm optimization–enhanced artificial neural network (PSO–ANN) that could predict the coconut milk spray drying process. The parameters for PSO tuning were selected as the number of particles and acceleration constant, respectively, for both global and personal best using a 2k factorial design. The optimal PSO settings were recorded as global best, C1 = 4.0; personal best, C2 = 0; and number of particles = 100. When comparing different types of spray drying models, PSO–ANN had an MSE value of 0.077, GA–ANN had an MSE of 0.033, while ANN had an MSE of 0.082. Sensitivity analysis was conducted on all three models to evaluate the significance level of each parameter on the model, and it was discovered that inlet temperature had the most significant influence on the model performance. In conclusion, the PSO–ANN was found to be more effective than ANN but less effective than GA–ANN in predicting the quality of coconut milk powder.

In this study, the development of an optimized topology neural network model for spray drying coconut milk is investigated using K-fold cross validation technique. Performance between standalone ANN and ANN with K-fold cross validation is compared, as K-fold cross validation method is integrated into neural network to overcome the limitations of restricted dataset. With inlet temperature (140 °C-180 °C), concentration of maltodextrin and sodium caseinate (0 w/w %- 10w/w %) are established as the input parameters, while moisture content (3.64%-5.1%), outlet temperature (76.5 °C-104.5 °C) and surface free fat percentage (0.35%-34.51%) are the output parameters for the neural network. Experimental data from the spray drying process is used to develop the neural network. Selection from the best training algorithm (gradient descent backpropagation, gradient descent with momentum, resilience backpropagation, conjugate gradient backpropagation with Polak-Riebre restarts, conjugate gradient backpropagation with Fletcher-Reeves, scaled conjugate gradient, Broyden-Flectcher-Goldfard-Shanno backpropagation algorithm and Levenberg-Marquardt backpropagation), transfer function (tansig, logsig, purelin and satlin), number of training runs (1000-5000), number of hidden layers (1-3) and nodes (5-15) have significant effect on the performance of the ANN models based on the lowest MSE values and R2 values. Overall, the optimum topology ANN model with k-fold cross validation outperformed the recorded lowest MSE value of 0.064 and highest R2 value of 0.855 compared to the optimum standalone ANN model with MSE value of 0.082 and R2 value of 0.832. The optimum ANN with K-fold cross validation implements the Levenberg-Marquart training algorithm with hyperbolic tangent sigmoid transfer function using 4500 times training runs with optimal topology configuration of 3-8-2-3. Result concludes that the developed neural network using K-fold cross validation represents the spray drying process as a highly reliable model with high degree of accuracy.

Application of Artificial Neural Network (ANN) and Genetic Algorithm (GA) are to provide an accurate model of the spray drying system. In this study, a comparative study is performed between ANN and GA enhanced ANN to estimate their abilities in emulating the spray drying process of coconut milk powder under restricted parameters. The GA parameter is optimized through response surface methodology (RSM). Through RSM, GA parameter such as population size, mutation and crossover are optimized and is used for the development of GA-ANN network. The optimized GA parameters values are at maximum population size (100), minimum crossover rate (0.2) and maximum mutation rate (1.0). The optimized GA parameters is then applied in the development of the GA-ANN network as the networks applied genetic algorithm to determine the initial weights in its neural network. Both models are then compared by placing importance of highest correlation of determination (R2) and lowest mean square error (MSE) values. The results have shown GA-ANN’s MSE (0.033396) is lower than ANN’s MSE value (0.082263), which the GA-ANN’s R2 value (0.88245) is higher than ANN (0.8499). This have shown that GA as a global search technique can be integrated in the development of the ANN.