Artificial neural networks for predicting the viscosity of lead-containing glasses

TL;DR

This paper develops an artificial neural network model to accurately predict the viscosity of lead-containing glasses based on chemical composition and temperature, outperforming existing models.

Contribution

It introduces a neural network approach with optimized node variation that improves viscosity prediction accuracy over previous models.

Findings

The best neural network model showed superior mean absolute error and R^2 compared to literature models.

Sensitivity analysis aligned with existing literature.

Predicted viscosity values correlated well with experimental data.

Abstract

The viscosity of lead-containing glasses is of fundamental importance for the manufacturing process, and can be predicted by algorithms such as artificial neural networks. The SciGlass database was used to provide training, validation and test data of chemical composition, temperature and viscosity for the construction of artificial neural networks with node variation in the hidden layer. The best model built with training data and validation data was compared with 7 other models from the literature, demonstrating better statistical evaluations of mean absolute error and coefficient of determination to the test data, with subsequent sensitivity analysis in agreement with the literature. Skewness and kurtosis were calculated and there is a good correlation between the values predicted by the best neural network built with the test data.