Interpretable Machine Learning Framework for Biomass–Plastic Co-gasification. (IMAGE)
Caption
Interpretable Machine Learning Framework for Biomass–Plastic Co-gasification. This graphical workflow illustrates the development of an interpretable machine learning framework to predict syngas composition during biomass–plastic co-gasification. The study integrates data collection from feedstock characteristics and operating conditions, preprocessing steps such as scaling and one-hot encoding, and model training using four algorithms. CatBoost demonstrated the highest predictive performance, and SHAP analysis was used to reveal the key factors influencing syngas yields, including temperature, steam-to-fuel ratio, and feedstock composition.
Credit
The authors
Usage Restrictions
None
License
Original content