Machine learning models and dataset for the prediction of Cr6+ removal of aqueous solutions using the pine cone residue

The experimental results of the adsorption of Cr6+ from aqueous solutions using the pine cone residue and the machine learning models used to investigate the experimental parameters are available on this page. Three adsorption conditions are optimized: contact time, pH, and initial solution concentration. The file “Pinha.csv” presents the results of the experiments considering various adsorption conditions. In addition, three machine learning models are employed to predict the behaviour of the experimental results: a multiple linear regression model, a decision tree model, and a random forest model. The machine learning models are in the “Machine learning models.ipynb” file. The maximum Cr6+ removal obtained with the pine cone residue is near 91%, and the machine learning models presented a high correlation coefficient of over 0.9, highlighting the potential of this type of methodology to enhance experimental studies.

本页面公开了利用松果残渣（pine cone residue）从水溶液中吸附六价铬（Cr6+）的实验结果，以及用于探究该实验参数的机器学习模型（machine learning models）相关资料。本次研究优化了三项核心吸附实验条件：接触时间、溶液pH值以及初始溶液浓度。文件"Pinha.csv"记录了不同吸附工况下的完整实验结果。此外，本研究采用三类机器学习模型对实验结果的变化规律进行预测：多元线性回归模型（multiple linear regression model）、决策树模型（decision tree model）与随机森林模型（random forest model）。上述机器学习模型的代码与运行结果均存放于"Machine learning models.ipynb"文件中。本研究中，采用松果残渣实现的六价铬最大去除率接近91%，且所构建的机器学习模型均取得了0.9以上的高相关系数（correlation coefficient），凸显了此类方法对优化实验研究的应用潜力。