This repository comprises all the code and data necessary to reproduce the results of the paper titled "The development of a design-oriented machine learning surrogate model for carbon capture with the implementation of explainable artificial intelligence."
This is an original work developed by researchers at LUT University. All rights reserved for LUT University.
- Make sure all necessary python libraries are installed: Numpy, Pandas, Scipy, Matplotlib, Seaborn, Scikit-learn, TensorFlow, and shap.
- Download the
data
folder, which contains the data used for this code. - Run the
makedirectories.ipynb
notebook to make the required directories. - Run each HPT notebook (
DT_HPT.ipynb
,RF_HPT.ipynb
,XG_HPT.ipynb
,SV_HPT.ipynb
, andDN_HPT.ipynb
) to see how the hyperparameters of each machine learning model are optimized. (Optional) - Run the
Main.ipynb
to reproduce the paper results.
Note that the size of the Main.ipynb notebook is too large to be rendered on GitHub; you can copy and paste its directory (https://github.com/Kasra-Aliyon/Surrogate-Machine-Learning-CCS/blob/main/Main.ipynb) to https://nbviewer.org/ and see it there. It will take some minutes to load.