Hybrid multicomponent UNIFAC density model of unsymmetrical dimethylhydrazine enhanced by machine learning and microscopic mechanism analysis

This study developed a machine learning bias-correction scheme for the UNIFAC model, incorporating temperature and key impurities to significantly improve UDMH density prediction accuracy within ±0.5 mg/cm³ error while maintaining interpretability. This dataset contains the machine-learing data and the quantum-chemical calculation data, if you need to obtain our data, please provide the user's name, institutional information, and a statement of purpose, and send it to the email address [e-mail] or xuzl@[e-mail].