Prediction of higher heating value of biomass materials based on proximate analysis using gradient boosted regression trees method

In the present research work, a machine learning tool based on the gradient boosted regression trees (GBRT) was used to predict the HHV of biomass. Data of 511 biomass samples were used to develop GBRT for prediction of HHV by utilizing proximate analysis. The values of mean absolute percentage error, root-mean-square error, and the determination coefficient for the developed model were 3.783%, 0.946, and 0.93, respectively, which represents high precision of HHV predictive capability. By comparing the models used to predict HHV, it was proved that the proposed model is better than the models found in literature so far.

Automated summary

In this research, a machine learning tool based on gradient boosted regression trees (GBRT) was used to predict the HHV of biomass. The developed model showed high precision in HHV prediction compared to previous models in the literature.