Optimization of genetic algorithm parameter in hybrid genetic algorithm-neural network modelling: Application to spray drying of coconut milk

Abstract

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.

Development of an Artificial Neural Network Utilizing Particle Swarm Optimization for Modeling the Spray Drying of Coconut Milk

“…PSO optimization techniques were integrated into the ANN to determine the optimum weights in the neural network design using MATLAB version 2019a. Lastly, the developed PSO-ANN was compared with external ANN [27] and GA-ANN [28] based on MSE and R 2 evaluation and supported with sensitivity analysis.…”

“…Based on the authors' previous studies, the neural network consisted of three input nodes and three output nodes with a topology configuration of 3-8-2-3. The development of the ANN was based on the Levenberg-Marquardt learning algorithm with a hyperbolic tangent sigmoid transfer function [28]. Based on the validation neural network results, the neural network design recorded a value of 0.064 for MSE and an R 2 value of 0.855 [37].…”

“…Figure 6 shows the results of MSE values and R 2 values tabulated using MATLAB. Comparisons were made between this PSO-ANN, GA-ANN [29], and ANN [28] using their MSE and R 2 values. The results showed that GA-ANN's MSE was the lowest (0.033) compared with the PSO-ANN value (0.077) and the ANN value (0.082).…”