A neural network approach to prediction of glass transition temperature

In today's world, there is a constant need to improve and optimise production processes and to create new materials with specific properties to meet the requirements of various industries. Glass, as a versatile material used in a multitude of fields, including construction, electronics, medicine and many others, is the object of increased attention. Many solutions exist, however, even with existing technologies, predicting the properties of glass based on its chemical composition remains a challenging task due to the complex relationship between the components and the physical characteristics of the material. Therefore, the development of accurate and efficient methods to address this problem is of high relevance and has the potential to significantly improve the production and utilisation of glass. A key step in the design of glass composition is the determination of its physical and chemical properties, such as the glass transition temperature (Tg).