Statistical comparison of models and machine learning for predicting Arrhenius parameters in ternary liquid mixtures

Viscosity analysis of fluid systems is crucial for optimizing industrial processes and products, especially when dealing with liquid mixtures. Numerous empirical models have been proposed to study the viscosity of pure, binary, and ternary liquid mixtures, including Hamdi's model. This model correlates two parameters from the Arrhenius-type viscosity equation, enabling the calculation of viscosity in both pure and binary liquid mixtures. This study extends the model to ternary liquid systems and compares its accuracy with other models. Additionally, we evaluate its predictive capability alongside machine learning models, such as linear regression, support vector regression, and random forest, in estimating Arrhenius parameters. A statistical analysis of 113 data points from previous studies was conducted for this comparison. The results demonstrate that Hamdi's model yields the highest accuracy, with predictions closely aligning with those of the machine learning models. Thus, the Hamdi model proves to be a valuable tool for industrial manufacturing.