Physics-informed machine learning for predicting temperature-dependent chemical properties

Key Points

• The aim is to develop a machine learning model to predict chemical properties based on temperature and chemical structure.
• Utilized physics-informed machine learning techniques
• Analyzed the relationship between chemical structure and thermodynamic properties
• Explored temperature dependencies in chemical properties
• Predictions closely matched experimental data
• Demonstrated improved accuracy over traditional models
• Highlighted significant temperature effects on properties

Abstract