Machine Learning Prediction of the 28-Day Compressive Strength of Ambient-Cured Geopolymer Concrete Using Ensemble Learning Algorithms
*,
DOI: 10.67571/njeru.2025.017481
Published: June 20, 2025 | Issue: Volume 2, Issue 1
Page No. 74–81
Read More
Email: mariabisong@unicross.edu.ng
1 Department of Mechanical Engineering, Faculty of Engineering, University of Cross River State, Calabar - Nigeria.
2 Department of Mechanical Engineering, Faculty of Engineering, University of Cross River State, Calabar - Nigeria.
Received: April 2, 2025 Reviewed: May 12, 2025 Accepted: June 20, 2025
Abstract
The nonlinear interactions among precursor composition, alkaline activator characteristics and mixture proportions complicate the prediction of compressive strength in geopolymer concrete. This study evaluated ensemble machine-learning algorithms for predicting the 28-day compressive strength of ambient-cured geopolymer concrete. Following data screening and removal of incomplete and duplicate observations, 251 mixtures were retained. Nine variables comprising fly ash, ground-granulated blast-furnace slag (GGBS), fine aggregate, coarse aggregate, NaOH, Na₂SiO₃, alkaline liquid-to-fly ash ratio, water-to-solids ratio and NaOH molarity were used as predictors. Random Forest, XGBoost, LightGBM, CatBoost and AdaBoost were evaluated using repeated five-fold cross-validation, with model performance assessed using the coefficient of determination (R2), root mean square error (RMSE) and mean absolute error (MAE). Correlation analysis showed that GGBS exhibited the strongest positive linear association with compressive strength (r = 0.388), while fly ash showed a moderate negative association (r = −0.427). CatBoost achieved the highest predictive performance, with a mean cross-validation R2 of 0.782 ± 0.100, RMSE of 8.13 ± 1.51 MPa and MAE of 5.37 ± 0.73 MPa, although XGBoost, LightGBM and Random Forest produced comparable performance. The results demonstrate that tree-based ensemble models can explain approximately 75–78% of the variation in 28-day compressive strength using routinely reported mixture and activator parameters. The models are best suited as preliminary screening tools within the represented mixture and ambient-curing domain rather than as substitutes for experimental validation.
- [1] D. Hardjito and B. V Rangan, “Development and Properties of Low Calcium Fly Ash-based Geopolymer Concrete,” Curtin Univ. Technol., 2005.
- [2] A. E. Kurtoğlu, R. Alzeebaree, O. Aljumaili, A. Niş, and A. Ç. Mehmet Eren Gülşan Ghassan Humur, “Mechanical and durability properties of fly ash and slag based geopolymer concrete,” Adv. Concr. Constr., vol. 4, pp. 345–362, 2018.
- [3] A. Mellado, C. Catalan, N. Bouzon, M. V. Borrachero, J. M. Monzo, and J. Paya, “Carbon footprint of geopolymeric mortar : study of and assessment of an alternative route,” pp. 23846–23852, 2014, doi: 10.1039/c4ra03375b.
- [4] A. Ahmad et al., “Prediction of Geopolymer Concrete Compressive Strength Using Novel Machine Learning Algorithms,” Polymers (Basel)., vol. 13, no. 19, p. 3389, 2021.
- [5] R. Cao, Z. Fang, M. Jin, and Y. Shang, “Application of Machine Learning Approaches to Predict the Strength Property of Geopolymer Concrete,” Materials (Basel)., vol. 15, no. 7, p. 2400, 2022.
- [6] Q. Wang, W. Ahmad, A. Ahmad, F. Aslam, A. Mohamed, and N. I. Vatin, “Application of Soft Computing Techniques to Predict the Strength of Geopolymer Composites,” Polymers (Basel)., vol. 14, no. 6, p. 1074, 2022.
- [7] A. Q. Khan, M. H. Naveed, M. D. Rasheed, and P. Miao, “Prediction of compressive strength of fly ash-based geopolymer concrete using supervised machine learning methods.,” Arab. J. Sci. Eng., vol. 49, no. 4, pp. 4889–4904, 2024, doi: 10.1007/s13369-023-08283-w.
- [8] K. Khan, W. Ahmad, M. N. Amin, A. Ahmad, S. Nazar, and M. A. Al-faiad, “Assessment of Artificial Intelligence Strategies to Estimate the Strength of Geopolymer Composites and Influence of Input Parameters,” Polymers (Basel)., vol. 14, no. 12, 2022.
- [9] G. S. Ryu, Y. B. Lee, K. T. Koh, and Y. S. Chung, “The mechanical properties of fly ash-based geopolymer concrete with alkaline activators,” Constr. Build. Mater., vol. 47, no. 2013, pp. 409–418, 2015.
- [10] S. R. Raj, P. G. Arulraj, N. Anand, K. Balamurali, and G. Gokul, “Influence of Various Design Parameters on Compressive Strength of Geopolymer Concrete : A Parametric Study by Taguchi Method,” Int. J. Eng., vol. 34, no. 10, pp. 2351–2359, 2021.
- [11] M. Olivia and H. Nikraz, “Properties of fly ash geopolymer concrete designed by Taguchi method,” Mater. Des., vol. 36, pp. 191–198, 2012, doi: 10.1016/j.matdes.2011.10.036.
- [12] S. V Dave and A. Bhogayata, “The strength oriented mix design for geopolymer concrete using Taguchi method and Indian concrete mix design code,” Constr. Build. Mater., vol. 262, p. 120853, 2020, doi: 10.1016/j.conbuildmat.2020.120853.
- [13] M. Anitha, A. Garg, and T. S. R. Babu, “Experimental study of geopolymer concrete with recycled fine aggregates and alkali activators,” Case Stud. Chem. Environ. Eng., vol. 8, p. 100501, 2023, doi: 10.1016/j.cscee.2023.100501.
- [14] G. James, D. Witten, T. Hastie, and R. Tibshirani, An Introduction to Statistical Learning: With applications in R, 2nd ed. Springer: Springer, 2021.
- [15] L. Breiman, “Random Forests,” Mach. Learn., vol. 45, no. 1, pp. 5–32, 2001.
- [16] Y. Freund and R. E. Schapire, “A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting *,” J. Comput. Syst. Sci., vol. 55, no. 1, pp. 119–139, 1997.
- [17] T. Chen and C. Guestrin, “XGBoost : A Scalable Tree Boosting System,” in Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, pp. 785–794.
- [18] G. Ke et al., “NIPS’17: Proceedings of the 31st International Conference on Neural Information Processing SystemsLightGBM: a highly efficient gradient boosting decision tree,” in NIPS’17: Proceedings of the 31st International Conference on Neural Information Processing Systems, 2017, pp. 3146–3154.
- [19] L. Prokhorenkova, G. Gusev, A. Vorobev, A. V. Dorogush, and A. Gulin, “CatBoost : unbiased boosting with categorical features,” Adv. Neural Inf. Process. Syst., vol. 31, no. Section 4, pp. 1–23, 2018.
- [20] T. Hastie, R. Tibshirani, and J. Friedman, The Elements of Statistical Learning: Data mining, inference and prediction, 2nd ed. Springer: Springer, 2009.
- [21] V. Q. Tran, “Data-driven approach for investigating and predicting of compressive strength of fly ash–slag geopolymer concrete,” Struct. Concr., vol. 24, no. 6, pp. 7419–7444, 2023.
