Interpretable machine-learning model with a collaborative game approach to predict yields and higher heating value of torrefied biomass

Machine learning models developed to predict energy properties of torrefied biomass. Collaborative game theory adopted to aid interpretability of key variables in torrefaction. Gradient boosting offered the highest prediction accuracy with 22-feature input. Novel framework to explain local and global effects of each feature on torrefaction. Torrefaction is a treatment process for converting biomass to high-quality solid fuels.