This study investigates the efficacy of feedforward neural network and XGBoost models in screening ionic liquid solvents for CO₂ capture. Both models were integrated with either group contribution (GC), molecular structure descriptors (MSD), or hybrid GC–MSD, to enable performance comparisons. It was demonstrated that the XGBoost models performed better over feedforward neural network models, irrespective of descriptor types. Notably, the XGBoost–GC–MSD model outperformed the artificial neural network with group contribution (ANN–GC) and structure encoding multilayer perceptron (SE-MLP) models from previous work, demonstrating an R² value of 0.98963, MAE of 0.01480, and RMSE of 0.02369. Even the least-performing XGBoost–GC model surpassed earlier ANN–GC and SE-MLP models, showcasing an R² value of 0.98891. Lastly, the Shapley additive explanation analysis identified the top five influential input features, including pressure, temperature, Chi2v, Chi0n, and BertzCT. These findings provide valuable insights into the molecular determinants affecting CO₂ solubility in ionic liquids.

中文翻译：

---

用于预测离子液体中 CO2 溶解度的可解释前馈神经网络和基于 XGBoost 的算法

_{本研究探讨了前馈神经网络和 XGBoost 模型在筛选用于 CO 2}捕获的离子液体溶剂中的功效。两种模型均与基团贡献 (GC)、分子结构描述符 (MSD) 或混合 GC-MSD 集成，以实现性能比较。事实证明，无论描述符类型如何，XGBoost 模型都比前馈神经网络模型表现更好。值得注意的是，XGBoost-GC-MSD 模型的性能优于先前工作中的群贡献人工神经网络 (ANN-GC) 和结构编码多层感知器 (SE-MLP) 模型，其R ²值为 0.98963，MAE 为 0.01480，并且RMSE 为 0.02369。即使是性能最差的 XGBoost-GC 模型也超越了早期的 ANN-GC 和 SE-MLP 模型，其R ²值为 0.98891。最后，Shapley 附加解释分析确定了影响最大的五个输入特征，包括压力、温度、Chi2v、Chi0n 和 BertzCT。这些发现为影响离子液体中CO ₂溶解度的分子决定因素提供了有价值的见解。