Prediction of reduced glass transition temperature of metallic alloys based on a neural network

Abstract

The reduced glass transition temperature Trg is an important glass forming ability parameter. Trg describes the glass formation in materials and the behaviour of materials at the transition between solid and liquid states and is an important parameter for materials analysis, development, and production process. This article describes the process and results of research on the development of a system for prediction of the reduced glass transition temperature Trg of metallic alloys based on recurrent neural network algorithms. The developed system can predict the reduced glass transition temperature Trg of metallic alloys based on the analysis of its chemical formula with high accuracy. The accuracy was evaluated using the 3 metrics: MSE, RMSE, MAE. Obtained values are: MSE value is 0.000678, RMSE value is 0.0260, MAE value is 0.01835.