Interpretable Ensemble Machine Learning for Liquefaction Risk Prediction

Tuzelbayev, Doszhan; Moon, Sung-Woo; Lee, Minho; Abdialim, Shynggys; Aremu, Elijah Adebayonle; Satyanaga, Alfrendo; Kim, Jong

doi:10.3390/infrastructures10110304

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

Interpretable Ensemble Machine Learning for Liquefaction Risk Prediction

by

Doszhan Tuzelbayev

¹,

Sung-Woo Moon

^1,*

,

Minho Lee

²

,

Shynggys Abdialim

¹,

Elijah Adebayonle Aremu

¹

,

Alfrendo Satyanaga

¹

and

Jong Kim

¹

Department of Civil and Environmental Engineering, Nazarbayev University, 53 Kabanbay Batyr Ave, Astana 010000, Kazakhstan

²

Department of Computer Science, Nazarbayev University, 53 Kabanbay Batyr Ave, Astana 010000, Kazakhstan

^*

Author to whom correspondence should be addressed.

Infrastructures 2025, 10(11), 304; https://doi.org/10.3390/infrastructures10110304

Submission received: 24 September 2025 / Revised: 4 November 2025 / Accepted: 7 November 2025 / Published: 11 November 2025

(This article belongs to the Special Issue Advances in Artificial Intelligence for Geotechnical Engineering)

Download Versions Notes

Abstract

This paper presents a comprehensive machine learning (ML) framework for predicting liquefaction risk, a crucial aspect of seismic hazard assessment. A benchmark geotechnical dataset with multi-dimensional input features was used to evaluate several ML classifiers, followed by hyperparameter optimization through stratified 5-fold cross-validation. Optimized models were combined into a soft Voting Ensemble to enhance stability and accuracy of liquefaction potential prediction. The proposed ensemble model achieved a mean accuracy of 90.12% and a recall of 97.23%, outperforming individual models in most folds. The ensemble’s effectiveness was further evidenced by its precision-recall (PR) and receiver operating characteristic (ROC) curves, with areas under the curve (AUC) of 0.962 and 0.931, respectively—closely matching those of the Gradient Boosting classifier, indicating comparable discriminatory performance. Additionally, SHapley Additive exPlanations (SHAP) analysis was conducted on the ensemble model to assess contributions of each geotechnical inputs to the predictions, revealing that normalized shear wave velocity (

V_{S 1}

) as the most influential variable in liquefaction prediction. The proposed framework demonstrates a robust, interpretable, and performance-consistent approach for liquefaction risk assessment.

Keywords: liquefaction; machine learning; ensemble learning; hyperparameter optimization; SHAP analysis

Share and Cite

MDPI and ACS Style

Tuzelbayev, D.; Moon, S.-W.; Lee, M.; Abdialim, S.; Aremu, E.A.; Satyanaga, A.; Kim, J. Interpretable Ensemble Machine Learning for Liquefaction Risk Prediction. Infrastructures 2025, 10, 304. https://doi.org/10.3390/infrastructures10110304

AMA Style

Tuzelbayev D, Moon S-W, Lee M, Abdialim S, Aremu EA, Satyanaga A, Kim J. Interpretable Ensemble Machine Learning for Liquefaction Risk Prediction. Infrastructures. 2025; 10(11):304. https://doi.org/10.3390/infrastructures10110304

Chicago/Turabian Style

Tuzelbayev, Doszhan, Sung-Woo Moon, Minho Lee, Shynggys Abdialim, Elijah Adebayonle Aremu, Alfrendo Satyanaga, and Jong Kim. 2025. "Interpretable Ensemble Machine Learning for Liquefaction Risk Prediction" Infrastructures 10, no. 11: 304. https://doi.org/10.3390/infrastructures10110304

APA Style

Tuzelbayev, D., Moon, S.-W., Lee, M., Abdialim, S., Aremu, E. A., Satyanaga, A., & Kim, J. (2025). Interpretable Ensemble Machine Learning for Liquefaction Risk Prediction. Infrastructures, 10(11), 304. https://doi.org/10.3390/infrastructures10110304

Article Menu

Interpretable Ensemble Machine Learning for Liquefaction Risk Prediction

Abstract

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI