Advanced Modeling Approaches for Quality Assessment of Papaya Leather Dried in IoT-Enabled IR-Assisted Refractance Window Dryer

Abstract

The aim of this research is the comprehensive comparison of advanced modeling tools for predicting the quality parameters of IoT-enabled IR-assisted RW-dried papaya leather. The models used in the current research were “response surface methodology (RSM),” “artificial neural network (ANN),” “machine learning regression algorithms (MLRA),” and “adaptive neuro-fuzzy inference system (ANFIS).” This comprehensive compression could be differentiated based on the model's fitness and prediction capabilities. The fitness of the model was assessed using statistical metrics such as “root mean square error (RMSE),” “mean square error (MAE),” and “coefficient of determination (R²).” While the prediction capability was evaluated by using tactics like predicting error, predicting accuracy, chi-square, and associated p-values. The results indicated that both the RSM and MLRA models had substantial predictive capabilities and achieved a prediction accuracy of 98.992 and 97.169, respectively. This study concluded that the model's fitness was getting excellent in the Alpha ANN (R² = 0.9943) and ANFIS (R² = 0.9903) models. However, when evaluating the prediction capabilities, the RSM and MLRA models outperformed the others.