Explainable Machine Learning Framework with Experimental Validation for Strength Prediction of Magnesium Phosphate Cement
Source abstract
Abstract
Abstract Magnesium Phosphate Cement (MPC) is recognized as an effective rapid repair material, with compressive strength serving as a key mechanical property indicator for its mortar formulations. Nevertheless, due to MPC's complex composition and formulation, predicting its compressive strength remains a significant challenge. In this study, a comprehensive database was developed, incorporating four key input variables: the magnesium-to-phosphate (M/P) molar ratio, water-to-cement (W/C) mass ratio, sand-to-binder (S/B) weight ratio, and the borax-to-magnesia(B/M) weight ratio. This dataset was used to train and validate eight machine learning models, including the Lightweight Gradient Boosting (LGB) algorithm, Support Vector Machine (SVM), Decision Tree (DT), Extreme Gradient Boosting (XGB), Ridge Regression (RR), Random Forest (RF), Backpropagation Neural Network (BP), and Gradient Boosting (GB) models. The eight machine learning models were evaluated using performance metrics, including Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), Correlation Coefficient, and Root Mean Square Error (RMSE), to identify the optimal model, which was then optimized via the Gray Wolf Optimizer (GWO). The most accurate prediction of MPC compressive strength was attained using the XGB model, with the GWO-optimized XGB model showing enhancement in MAPE, MAE, R2, and RMSE by 21.8%, 60.6%, 43.9%, and 55.3% respectively, relative to the unoptimized XGB model. Employing Shapley Additive exPlanations (SHAP) values and Partial Dependence Plots (PDP), this study facilitates the identification of the most influential input variables and quantifies their effects on MPC compressive strength. The optimized model was validated against experimental data, demonstrating robust and conservative prediction behavior. While the model is trained solely to predict compressive strength, its interpretability enables rational insights into how formulation variables influence strength, thereby supporting informed mix design decisions. This framework offers a reliable and transparent computational tool for preemptive strength assessment of MPC and guides the optimization of mechanical performance in structurally demanding applications.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Magnesium Oxide Properties and Applications · Magnesium Alloys: Properties and Applications · Innovative concrete reinforcement materials
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (39)
Show source-provided references
1. Abdelrazig, B. E. I., Sharp, J. H., & El-Jazairi, B. (1988). The chemical composition of mortars made from magnesia-phosphate cement. Cement and Concrete Research, 18(3), 415–425. https://doi.org/10.1016/0008-8846(88)90075-010.1016/0008-8846(88)90075-0 ↗
2. Ahmed, S. R., Ahmed, A. K., & Jwmaa, S. J. (2023). Analyzing the employee turnover by using decision tree algorithm. In 2023 5th international congress on human-computer interaction, optimization and robotic applications (HORA) (pp. 1–4). IEEE. https://doi.org/10.1109/HORA58378.2023.1015670910.1109/HORA58378.2023.10156709 ↗
3. Chau, C. K., Qiao, F., & Li, Z. (2011). Microstructure of magnesium potassium phosphate cement. Construction and Building Materials, 25(6), 2911–2917. https://doi.org/10.1016/j.conbuildmat.2010.12.03510.1016/j.conbuildmat.2010.12.035 ↗
4. Chen, T., & Guestrin, C. (2016). Xgboost: A scalable tree boosting system. In: Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining (pp. 785–794). https://doi.org/10.1145/2939672.293978510.1145/2939672.2939785 ↗
5. Delgado, J., Silva, F. A. N., Azevedo, A. C., et al. (2020). Artificial neural networks to assess the useful life of reinforced concrete elements deteriorated by accelerated chloride tests. Journal of Building Engineering, 31, Article 101445. https://doi.org/10.1016/j.jobe.2020.10144510.1016/j.jobe.2020.101445 ↗
6. Ding, Z., & Li, Z. (2005). High-early-strength magnesium phosphate cement with fly ash. ACI Materials Journal, 102(6), 375.
7. Friedman, J. H. (2002). Stochastic gradient boosting. Computational Statistics & Data Analysis, 38(4), 367–378. https://doi.org/10.1016/S0167-9473(01)00065-210.1016/S0167-9473(01)00065-2 ↗
8. Guo, X., Liu, X., Long, G., et al. (2023). Data-driven prediction of the fatigue performance of corroded high-strength steel wires. Engineering Failure Analysis, 146, Article 107108. https://doi.org/10.1016/j.engfailanal.2023.10710810.1016/j.engfailanal.2023.107108 ↗
9. Hou, D., Yan, H., Zhang, J., et al. (2016). Experimental and computational investigation of magnesium phosphate cement mortar. Construction and Building Materials, 112, 331–342. https://doi.org/10.1016/j.conbuildmat.2016.02.20010.1016/j.conbuildmat.2016.02.200 ↗
10. Khodadadi, N., Roghani, H., Harati, E., Mirdarsoltany, M., De Caso, F., & Nanni, A. (2024). Fiber-reinforced polymer (FRP) in concrete: A comprehensive survey. Construction and Building Materials, 432, 136634.https://doi.org/10.1016/j.conbuildmat.2024.136634.10.1016/j.conbuildmat.2024.136634 ↗
11. Le Rouzic, M., Chaussadent, T., Stefan, L., et al. (2017). On the influence of Mg/P ratio on the properties and durability of magnesium potassium phosphate cement pastes. Cement and Concrete Research, 96, 27–41. https://doi.org/10.1016/j.cemconres.2017.02.03310.1016/j.cemconres.2017.02.033 ↗
12. Lee, K. H., Yoon, H. S., & Yang, K. H. (2017). Tests on magnesium potassium phosphate composite mortars with different water-to-binder ratios and molar ratios of magnesium-to-phosphate. Construction and Building Materials, 146, 303–311. https://doi.org/10.1016/j.conbuildmat.2017.04.12310.1016/j.conbuildmat.2017.04.123 ↗
13. Li, Y., Sun, J., & Chen, B. (2014). Experimental study of magnesia and M/P ratio influencing properties of magnesium phosphate cement. Construction and Building Materials, 65, 177–183. https://doi.org/10.1016/j.conbuildmat.2014.04.13610.1016/j.conbuildmat.2014.04.136 ↗
14. Long, S., Li, Y., Wang, N., et al. (2024). Optimization of magnesium phosphate cement prepared by natural brucite using ultrafine metakaolin and metakaolin. Journal of Building Engineering, 98, Article 111459. https://doi.org/10.1016/j.jobe.2024.11145910.1016/j.jobe.2024.111459 ↗
15. Luo, X., Li, Y., Wang, Q., et al. (2024). Machine learning based modeling for predicting the compressive strength of solid waste material-incorporated Magnesium Phosphate Cement. Journal of Cleaner Production, 442, Article 141172. https://doi.org/10.1016/j.jclepro.2024.14117210.1016/j.jclepro.2024.141172 ↗
16. Ma, C., Wang, F., Zhou, H., et al. (2021). Effect of early-hydration behavior on rheological properties of borax-admixed magnesium phosphate cement. Construction and Building Materials, 283, Article 122701. https://doi.org/10.1016/j.conbuildmat.2021.12270110.1016/j.conbuildmat.2021.122701 ↗
17. Ma, H., Xu, B., Liu, J., et al. (2014). Effects of water content, magnesia-to-phosphate molar ratio and age on pore structure, strength and permeability of magnesium potassium phosphate cement paste. Materials & Design, 64, 497–502. https://doi.org/10.1016/j.matdes.2014.07.07310.1016/j.matdes.2014.07.073 ↗
18. Ma, S., Gu, J., Wang, J., et al. (2025). In-depth insight into the driving factors of the compressive strength development of MKPC based on interpretable machine learning methods. Journal of Industrial and Engineering Chemistry, 141, 305–318. https://doi.org/10.1016/j.jiec.2024.06.03910.1016/j.jiec.2024.06.039 ↗
19. Mangalathu, S., Heo, G., & Jeon, J. S. (2018). Artificial neural network based multi-dimensional fragility development of skewed concrete bridge classes. Engineering Structures, 162, 166–176. https://doi.org/10.1016/j.engstruct.2018.01.05310.1016/j.engstruct.2018.01.053 ↗
20. Mirjalili, S., Mirjalili, S. M., & Lewis, A. (2014). Grey wolf optimizer. Advances in engineering software, 69, 46-61. https://doi.org/10.1016/j.advengsoft.2013.12.00710.1016/j.advengsoft.2013.12.007 ↗
21. Natekin, A., & Knoll, A. (2013). Gradient boosting machines, a tutorial. Frontiers in Neurorobotics, 7, Article 21. https://doi.org/10.3389/fnbot.2013.0002110.3389/fnbot.2013.00021 ↗
22. Nehdi, M., Djebbar, Y., & Khan, A. (2001). Neural network model for preformed-foam cellular concrete. Materials Journal, 98(5), 402–409.
23. Pal, M. (2005). Random forest classifier for remote sensing classification. International Journal of Remote Sensing, 26(1), 217–222. https://doi.org/10.1080/0143116041233126969810.1080/01431160412331269698 ↗
24. Precup, R. E., David, R. C., & Petriu, E. M. (2016). Grey wolf optimizer algorithm-based tuning of fuzzy control systems with reduced parametric sensitivity. IEEE Transactions on Industrial Electronics, 64(1), 527–534. https://doi.org/10.1109/TIE.2016.260769810.1109/TIE.2016.2607698 ↗
25. Qiao, F., Chau, C. K., & Li, Z. (2010). Property evaluation of magnesium phosphate cement mortar as patch repair material. Construction and Building Materials, 24(5), 695–700. https://doi.org/10.1016/j.conbuildmat.2009.10.03910.1016/j.conbuildmat.2009.10.039 ↗
26. Qin, J., Qian, J., Dai, X., et al. (2021). Effect of water content on microstructure and properties of magnesium potassium phosphate cement pastes with different magnesia-to-phosphate ratios. Journal of the American Ceramic Society, 104(6), 2799–2819. https://doi.org/10.1111/jace.1769510.1111/jace.17695 ↗
27. Robertson, I., Masson, C., Sedran, T., et al. (2015). Advantages of a new ballastless trackform. Construction and Building Materials, 92, 16–22. https://doi.org/10.1016/j.conbuildmat.2014.06.09910.1016/j.conbuildmat.2014.06.099 ↗
28. Vapnik, V. (2000). The Nature of Statistical Learning Theory, 2nd ed. New York: Springer- Verlag. https://doi.org/10.1007/978-1-4757-3264-1.10.1007/978-1-4757-3264-1 ↗
29. Wang, X., Wang, X., Ma, B., et al. (2023). High-performance reversible data hiding based on ridge regression prediction algorithm. Signal Processing, 204, Article 108818. https://doi.org/10.1016/j.sigpro.2022.10881810.1016/j.sigpro.2022.108818 ↗
30. Xu, B., Ma, H., & Li, Z. (2015). Influence of magnesia-to-phosphate molar ratio on microstructures, mechanical properties and thermal conductivity of magnesium potassium phosphate cement paste with large water-to-solid ratio. Cement and Concrete Research, 68, 1–9. https://doi.org/10.1016/j.cemconres.2014.10.01910.1016/j.cemconres.2014.10.019 ↗
31. Yang, Q., & Wu, X. (1999). Factors influencing properties of phosphate cement-based binder for rapid repair of concrete. Cement and Concrete Research, 29(3), 389–396. https://doi.org/10.1016/S0008-8846(98)00230-010.1016/S0008-8846(98)00230-0 ↗
32. Yao, L., Ren, L., Gong, G., IOP Publishing. (2021). Evaluation of chloride diffusion in concrete using PSO-BP and BP neural network. IOP Conference Series: Earth and Environmental Science, 687(1), Article 012037. https://doi.org/10.1088/1755-1315/687/1/01203710.1088/1755-1315/687/1/012037 ↗
33. Yao, Y., Zhang, J., Yao, Z., et al. (2024). Evaluation of mechanical and microscopic properties of magnesium phosphate cement-based concrete with different fly ash additions. Journal of Building Engineering, 91, Article 109640. https://doi.org/10.1016/j.jobe.2024.10964010.1016/j.jobe.2024.109640 ↗
34. Yaseen, Z. M., Deo, R. C., Hilal, A., et al. (2018). Predicting compressive strength of lightweight foamed concrete using extreme learning machine model. Advances in Engineering Software, 115, 112–125. https://doi.org/10.1016/j.advengsoft.2017.09.00410.1016/j.advengsoft.2017.09.004 ↗
35. You, C., Qian, J., Qin, J., et al. (2015). Effect of early hydration temperature on hydration product and strength development of magnesium phosphate cement (MPC). Cement and Concrete Research, 78, 179–189. https://doi.org/10.1016/j.cemconres.2015.07.00510.1016/j.cemconres.2015.07.005 ↗
36. Yu, X., Hu, T., Khodadadi, N., et al. (2025a). Modeling chloride ion diffusion in recycled aggregate concrete: A fuzzy neural network approach integrating material and environmental factors. Structures, 73, Article 108372. https://doi.org/10.1016/j.istruc.2025.10837210.1016/j.istruc.2025.108372 ↗
37. Yu, X. R., Song, A., Khodadadi, N., et al. (2025b). A novel approach for predicting FRP debonding strain in concrete using an optimized self-learning model. Composite Structures. https://doi.org/10.1016/j.compstruct.2025.11957210.1016/j.compstruct.2025.119572 ↗
38. Zhang, H., Zhang, Q., Tang, S., et al. (2024a). Effects of fly ash and silica fume on the rheological properties of magnesium phosphate cement-emulsified asphalt (MPC-EA) composite repair materials. Construction and Building Materials, 433, Article 136708. https://doi.org/10.1016/j.conbuildmat.2024.13670810.1016/j.conbuildmat.2024.136708 ↗
39. Zhang, J., Li, T., Yao, Y., et al. (2024b). Optimization of mix proportion and strength prediction of magnesium phosphate cement-based composites based on machine learning. Construction and Building Materials, 411, Article 134738. https://doi.org/10.1016/j.conbuildmat.2023.13473810.1016/j.conbuildmat.2023.134738 ↗