Machine Learning and Explainable Artificial Intelligence for Predicting Unconfined Compressive Strength in Carbonated Stabilized Soils

Mohammed, Ahmed Mohammed Awad, Husain, Omayma, Salih, Sayeed, Hamdan Mohamed, Mosab and et al., - (2026) Machine Learning and Explainable Artificial Intelligence for Predicting Unconfined Compressive Strength in Carbonated Stabilized Soils. Transportation Infrastructure Geotechnology, 13 (7). ISSN 21967202

Full text not available from this repository.

Abstract

Machine learning approach based on explainable artificial intelligence was used for the prediction of the unconfined compressive strength (UCS) of carbonated soils using magnesium-containing stabilizers. A dataset of 170 samples was compiled, covering stabilizer content, magnesium oxide, calcium oxide, moisture content, dry density, curing period, carbonation period, and carbon dioxide pressure. Six models, namely Extreme Gradient Boosting (XGB), Gradient Boosting, Artificial Neural Network, Random Forest, Decision Tree, and Linear Regression, were trained and compared using the coefficient of determination (R²), root mean squared error (RMSE), and mean absolute error (MAE). XGB achieved the highest accuracy in predicting UCS, with an R² of 0.891, an RMSE of 0.396, and an MAE of 0.27, outperforming Gradient Boosting (R² = 0.858) and Decision Tree (R² = 0.855). Linear Regression recorded the weakest performance, with an R² of 0.56. Shapley Additive Explanations and Local Interpretable Model-agnostic Explanations applied to the XGB model identified stabilizer content and moisture content as the most influential variables governing UCS: moisture exerted a negative effect, while stabilizer content exerted a positive effect. Carbonation period and magnesium oxide showed moderate contributions. The model offered a reliable, low-cost alternative to laboratory testing and supported optimization of carbonation-based soil stabilization.

Item Type:	Article
Uncontrolled Keywords:	Explainable artificial intelligence; Machine learning; Mineral carbonation; SHAP analysis; Soil stabilization; Unconfined compressive strength
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning G Geography. Anthropology. Recreation > GE Environmental Sciences > Earth sciences > Geology > Physical geology > Sedimentation and deposition
Divisions:	School of Computing > Staff Research and Publications
Depositing User:	Tamara Malone
Date Deposited:	16 Sep 2026 17:57
Last Modified:	16 Sep 2026 17:57
URI:	https://norma.ncirl.ie/id/eprint/9931

Actions (login required)

View Item