Answers are grounded in this paper and cite their sources.
This study investigates the prediction of compressive strength (CS) of concrete incorporating biomass combustion fly ash (BCFA), a promising supplementary cementitious material for sustainable construction. This study uses a total of 359 concrete mixes along with their respective CS values collected from different sources. Eleven machine learning (ML) algorithms, including Decision Tree (DT), Random Forest (RF), Gradient Boosting Regressor (GBR), Light Gradient Boosting (LGB), Extreme Gradient Boosting (XGB), CatBoost (CatB), Histogram Gradient Boosting (HGB), AdaBoost (ADB), Bagging Regressor (BAG), K-Nearest Neighbors (KNN), and Multilayer Perceptron (MLP), have been developed and systematically evaluated using training and testing datasets. The results demonstrated that ensemble-based boosting models generally provided superior predictive performance. In the training dataset, DT, CatB, and XGB achieved the highest R² values of 0.99, while RF and GBR attained 0.97For the test set, the highest predictive ability was achieved by CatB model with an R² of 0.96, followed by GBR, HGB, and LGB models with R² of 0.94. SHapley Additive exPlanation (SHAP) and Partial Dependence Plot (PDP) analysis results indicated that the important input factors for CS prediction include biomass combustion fly ash (BCFA), water content, and curing age. Optimum range values of the input factors for strength development have been calculated, and GUI for quick CS prediction has been created. The findings show that accurate, comprehensible, and useful ecological concrete design using BCFA is made possible by combining explainable artificial intelligence with high-performing ML techniques.
No takes yet. Share an insight, caveat, or question.
Sobuz et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: