This study investigates the improvement in the compressive strength of cellulose/cement-based composites. Methyl cellulose (MC), carboxymethyl cellulose (CMC), and hydroxypropyl cellulose (HPMC) are separately used as the cellulose phase with different wt%. Graphene oxide (GO) and zoledronic acid (ZOL) are used as additives for bone regeneration for various formulations. Utilizing Extreme Gradient Boosting (XGB) modeling, this research demonstrates the roles of the choice of the cellulose phase, wt% of cement phase, % gelatin, % citric acid, degradation time, and concentration of GO and ZOL in influencing compressive strength. The XGB regression model, with an R2 value of 0.99 (~1), shows the predictive power of the model. Feature importance analysis demonstrates the significance of cellulose choice and the addition of chitosan in enhancing compressive strength. The correlation heatmap reveals positive associations, emphasizing the positive influence of HPMC and CMC compared with MC and the substantial impact of chitosan and citric acid on compressive strength. The model's predictive accuracy is validated through predicted compressive strength values with experimental observations, providing insights for optimizing cellulose-reinforced cements and enabling tailored material design for enhanced mechanical performance.
Read full abstract