Interpreting machine learning predictions of Pb2+ adsorption onto biochars produced by a fluidized bed system

作者：Suya Shi, Yaji Huang, Han‐Ming Shen, Tengfei Zheng, Xinye Wang, Mengzhu Yu, Lingqin Liu · 发表于：Journal of Cleaner Production · 年份：2024 · DOI：10.1016/j.jclepro.2024.144551 · 被引用次数：23 · 研究领域：Adsorption and biosorption for pollutant removal、Mineral Processing and Grinding

Employing machine learning to predict the Pb 2+ adsorption capacity of biochars is an innovative pursuit in hazardous materials. This study compared artificial neural network (ANN), support vector regression (SVR) and random forest (RF) for Pb 2+ adsorption capacity by biochar from a fluidized bed system. Besides developing correlations for comparison, the RF model (R 2 = 0.984, RMSE = 0.054) outperformed both ANN (R 2 = 0.908, RMSE = 0.316) and SVR (R 2 = 0.667) in predicting higher adsorption capacity. Based on the superior performance, the Shapley Additive Explanations (SHAP) were employed on RF. SHAP global explanations indicated that adsorption conditions contributed 69.03% and biochar characteristics contributed 30.21%to adsorption capacity, highlighting Dosage (D) and Carbon (C) as the crucial factors. Regarding biochar characteristics, element compositions contributed 76.59%. The single samples demonstrated that the final predictions align with the experimental results. The synergistic effect of dependence plot explains the Pb 2+ adsorption under varying parameter conditions, such as D < 1 g/L, C<45%, Pb in >100 mg/L, H < 2.5, t > 12h, T > 25 °C, pH > 9, H/C > 0.4, the SHAP value is positive, contributing to an increase in adsorption capacity. Furthermore, a graphical user interface (GUI) leveraging SHAP model parameters predicts adsorbent performance, providing novel insights into optimizing biochars production. The obtained findings narrow the search for optimal bio...