Machine Learning Prediction of Henry’s Law Constant for CO2 in Ionic Liquids and Deep Eutectic Solvents DOI Creative Commons
Dmitriy M. Makarov, Yu. A. Fadeeva, A. M. Kolker

и другие.

Liquids, Год журнала: 2025, Номер 5(2), С. 16 - 16

Опубликована: Май 30, 2025

Ionic liquids (ILs) and deep eutectic solvents (DESs) have been extensively studied as absorbents for CO2 capture, demonstrating high efficiency in this role. To optimize the search compounds with superior absorption properties, theoretical approaches, including machine learning methods, are highly relevant. In study, models were developed applied to predict Henry’s law constants ILs DESs, aiming identify systems best performance. The accuracy of was assessed interpolation tasks within training set extrapolation beyond its domain. optimal predictive built using CatBoost algorithm, leveraging CDK molecular descriptors RDKit DESs. define applicability domain models, SHAP-based leverage method employed, providing a quantitative characterization descriptor space where predictions remain reliable. integrated into web platform chem-predictor, they can be utilized predicting properties.

Язык: Английский

Machine Learning Prediction of Henry’s Law Constant for CO2 in Ionic Liquids and Deep Eutectic Solvents DOI Creative Commons
Dmitriy M. Makarov, Yu. A. Fadeeva, A. M. Kolker

и другие.

Liquids, Год журнала: 2025, Номер 5(2), С. 16 - 16

Опубликована: Май 30, 2025

Ionic liquids (ILs) and deep eutectic solvents (DESs) have been extensively studied as absorbents for CO2 capture, demonstrating high efficiency in this role. To optimize the search compounds with superior absorption properties, theoretical approaches, including machine learning methods, are highly relevant. In study, models were developed applied to predict Henry’s law constants ILs DESs, aiming identify systems best performance. The accuracy of was assessed interpolation tasks within training set extrapolation beyond its domain. optimal predictive built using CatBoost algorithm, leveraging CDK molecular descriptors RDKit DESs. define applicability domain models, SHAP-based leverage method employed, providing a quantitative characterization descriptor space where predictions remain reliable. integrated into web platform chem-predictor, they can be utilized predicting properties.

Язык: Английский

Процитировано

0