Active learning for predicting the enthalpy of mixing in binary liquids based on ab initio molecular dynamics

Key Points

• The aim is to develop a predictive model for the enthalpy of mixing in binary liquids using active learning techniques.
• Utilized ab initio molecular dynamics simulations
• Applied active learning algorithms to predict enthalpy
• Evaluated the model with various binary liquid mixtures
• Predictions demonstrate high accuracy compared to experimental values
• Active learning reduced the number of necessary simulations
• Insights gained into the molecular interactions affecting mixing