Development of the Optuna-NGBoost-SHAP model for estimating ground settlement during tunnel excavation

This study aims to develop and evaluate a natural gradient boosting (NGBoost) model optimized with Optuna for estimating ground settlement during tunnel excavation, incorporating Shapley additive explanations (SHAP) to perform interpretability analysis on the model’s estimation results. The model’s...

Full description

Saved in:
Bibliographic Details
Main Authors: Yuxin Chen, Mohammad Hossein Kadkhodaei, Jian Zhou
Format: Article
Language:English
Published: KeAi Communications Co., Ltd. 2025-10-01
Series:Underground Space
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2467967425000595
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849713280478609408
author Yuxin Chen
Mohammad Hossein Kadkhodaei
Jian Zhou
author_facet Yuxin Chen
Mohammad Hossein Kadkhodaei
Jian Zhou
author_sort Yuxin Chen
collection DOAJ
description This study aims to develop and evaluate a natural gradient boosting (NGBoost) model optimized with Optuna for estimating ground settlement during tunnel excavation, incorporating Shapley additive explanations (SHAP) to perform interpretability analysis on the model’s estimation results. The model’s predictive performance was comprehensively assessed using datasets from two earth pressure balance shield tunneling projects in Changsha and Zhengzhou, China. Comparative analyses demonstrated the superior accuracy and generalization capability of the Optuna-NGBoost-SHAP model (training set: R2 = 0.9984, MAE = 0.1004, RMSE = 0.4193, MedAE = 0.0122; validation set: R2 = 0.9001, MAE = 1.3363, RMSE = 3.2992, MedAE = 0.3042; test set: R2 = 0.9361, MAE = 0.9961, RMSE = 2.5388, MedAE = 0.2147). SHAP value analysis quantitatively evaluated the contributions of input features to the model’s estimations, identifying geometric factors (distance from the shield machine to the monitoring section and cover depth) as the most important features. The findings provide robust decision support for safety management during tunnel construction and demonstrate the reliability and efficiency of the Optuna-NGBoost-SHAP framework in estimating complex ground settlement scenarios.
format Article
id doaj-art-36bd378bb5e244cd846569f2cb5532d9
institution DOAJ
issn 2467-9674
language English
publishDate 2025-10-01
publisher KeAi Communications Co., Ltd.
record_format Article
series Underground Space
spelling doaj-art-36bd378bb5e244cd846569f2cb5532d92025-08-20T03:13:59ZengKeAi Communications Co., Ltd.Underground Space2467-96742025-10-0124607810.1016/j.undsp.2025.03.006Development of the Optuna-NGBoost-SHAP model for estimating ground settlement during tunnel excavationYuxin Chen0Mohammad Hossein Kadkhodaei1Jian Zhou2School of Resources and Safety Engineering, Central South University, Changsha 410083, ChinaIsfahan University of Technology, Department of Mining Engineering, Isfahan 8415683111, Iran; Corresponding author.School of Resources and Safety Engineering, Central South University, Changsha 410083, ChinaThis study aims to develop and evaluate a natural gradient boosting (NGBoost) model optimized with Optuna for estimating ground settlement during tunnel excavation, incorporating Shapley additive explanations (SHAP) to perform interpretability analysis on the model’s estimation results. The model’s predictive performance was comprehensively assessed using datasets from two earth pressure balance shield tunneling projects in Changsha and Zhengzhou, China. Comparative analyses demonstrated the superior accuracy and generalization capability of the Optuna-NGBoost-SHAP model (training set: R2 = 0.9984, MAE = 0.1004, RMSE = 0.4193, MedAE = 0.0122; validation set: R2 = 0.9001, MAE = 1.3363, RMSE = 3.2992, MedAE = 0.3042; test set: R2 = 0.9361, MAE = 0.9961, RMSE = 2.5388, MedAE = 0.2147). SHAP value analysis quantitatively evaluated the contributions of input features to the model’s estimations, identifying geometric factors (distance from the shield machine to the monitoring section and cover depth) as the most important features. The findings provide robust decision support for safety management during tunnel construction and demonstrate the reliability and efficiency of the Optuna-NGBoost-SHAP framework in estimating complex ground settlement scenarios.http://www.sciencedirect.com/science/article/pii/S2467967425000595Ground settlementHyperparameter optimizationNGBoostEstimationInterpretable machine learning
spellingShingle Yuxin Chen
Mohammad Hossein Kadkhodaei
Jian Zhou
Development of the Optuna-NGBoost-SHAP model for estimating ground settlement during tunnel excavation
Underground Space
Ground settlement
Hyperparameter optimization
NGBoost
Estimation
Interpretable machine learning
title Development of the Optuna-NGBoost-SHAP model for estimating ground settlement during tunnel excavation
title_full Development of the Optuna-NGBoost-SHAP model for estimating ground settlement during tunnel excavation
title_fullStr Development of the Optuna-NGBoost-SHAP model for estimating ground settlement during tunnel excavation
title_full_unstemmed Development of the Optuna-NGBoost-SHAP model for estimating ground settlement during tunnel excavation
title_short Development of the Optuna-NGBoost-SHAP model for estimating ground settlement during tunnel excavation
title_sort development of the optuna ngboost shap model for estimating ground settlement during tunnel excavation
topic Ground settlement
Hyperparameter optimization
NGBoost
Estimation
Interpretable machine learning
url http://www.sciencedirect.com/science/article/pii/S2467967425000595
work_keys_str_mv AT yuxinchen developmentoftheoptunangboostshapmodelforestimatinggroundsettlementduringtunnelexcavation
AT mohammadhosseinkadkhodaei developmentoftheoptunangboostshapmodelforestimatinggroundsettlementduringtunnelexcavation
AT jianzhou developmentoftheoptunangboostshapmodelforestimatinggroundsettlementduringtunnelexcavation