Yanjiang He, Ao Yang, Changjun Zou, Tianyou Fan, Qikui Lan, Yu He, Meng Wang, Jaka Sunarso, Zong Yang Kong
{"title":"An interpretable surrogate model for H2S solubility forecasting in ionic liquids based on machine learning","authors":"Yanjiang He, Ao Yang, Changjun Zou, Tianyou Fan, Qikui Lan, Yu He, Meng Wang, Jaka Sunarso, Zong Yang Kong","doi":"10.1016/j.seppur.2024.130061","DOIUrl":null,"url":null,"abstract":"Here we investigated four different ML-based models, i.e., gaussian process regression (GPR), extreme gradient boosting (i.e., XGBoost), random forest (RF), and support vector machine (SVM), for predicting the solubility of H<sub>2</sub>S in various ionic liquids (ILs). The dataset was divided into training and testing sets in an 80:20 ratio while the model performance for all models were evaluated using the coefficient of determination (R<sup>2</sup>), mean absolute error (MAE), and root mean square error (RMSE). Overall, all models effectively predicted H<sub>2</sub>S solubility, albeit with varying degrees of performance. The GPR provides the best performance, with R<sup>2</sup> of 0.9918, MAE of 0.0090, and RMSE of 0.0147. Following this is the XGBoost model with an R<sup>2</sup> value of 0.9827, MAE of 0.0155, and RMSE of 0.0213. The RF model displayed slightly lower performance, with an R<sup>2</sup> value of 0.9395, MAE of 0.0261, and RMSE of 0.0398 while the lowest performance was demonstrated by the SVM model, which gave an R<sup>2</sup> value of 0.9036, MAE of 0.0402, and RMSE of 0.0508. We used SHAP analysis, identified pressure, temperature, Estate_VSA3, Estate_VSA5, and MinEStateIndex as the top five dominant input features in our model interpretation. In a nutshell, this work presents new insights into the molecular characteristics that affect the solubility of H<sub>2</sub>S in ILs, paving future research path in this field.","PeriodicalId":427,"journal":{"name":"Separation and Purification Technology","volume":null,"pages":null},"PeriodicalIF":8.1000,"publicationDate":"2024-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Separation and Purification Technology","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1016/j.seppur.2024.130061","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, CHEMICAL","Score":null,"Total":0}
引用次数: 0
Abstract
Here we investigated four different ML-based models, i.e., gaussian process regression (GPR), extreme gradient boosting (i.e., XGBoost), random forest (RF), and support vector machine (SVM), for predicting the solubility of H2S in various ionic liquids (ILs). The dataset was divided into training and testing sets in an 80:20 ratio while the model performance for all models were evaluated using the coefficient of determination (R2), mean absolute error (MAE), and root mean square error (RMSE). Overall, all models effectively predicted H2S solubility, albeit with varying degrees of performance. The GPR provides the best performance, with R2 of 0.9918, MAE of 0.0090, and RMSE of 0.0147. Following this is the XGBoost model with an R2 value of 0.9827, MAE of 0.0155, and RMSE of 0.0213. The RF model displayed slightly lower performance, with an R2 value of 0.9395, MAE of 0.0261, and RMSE of 0.0398 while the lowest performance was demonstrated by the SVM model, which gave an R2 value of 0.9036, MAE of 0.0402, and RMSE of 0.0508. We used SHAP analysis, identified pressure, temperature, Estate_VSA3, Estate_VSA5, and MinEStateIndex as the top five dominant input features in our model interpretation. In a nutshell, this work presents new insights into the molecular characteristics that affect the solubility of H2S in ILs, paving future research path in this field.
期刊介绍:
Separation and Purification Technology is a premier journal committed to sharing innovative methods for separation and purification in chemical and environmental engineering, encompassing both homogeneous solutions and heterogeneous mixtures. Our scope includes the separation and/or purification of liquids, vapors, and gases, as well as carbon capture and separation techniques. However, it's important to note that methods solely intended for analytical purposes are not within the scope of the journal. Additionally, disciplines such as soil science, polymer science, and metallurgy fall outside the purview of Separation and Purification Technology. Join us in advancing the field of separation and purification methods for sustainable solutions in chemical and environmental engineering.