Full Text:
<44>
CLC number:
On-line Access: 2025-06-06
Received: 2024-12-10
Revision Accepted: 2025-03-26
Crosschecked: 0000-00-00
Cited: 0
Clicked: 65
Machine learning for soil parameter inversion enhanced with Bayesian optimization
| |||||||||||||
Anfeng HU1, Chi WANG1,2, Senlin XIE1, Zhirong XIAO3, Tang LI1, Ang XU1. Machine learning for soil parameter inversion enhanced with Bayesian optimization[J]. Journal of Zhejiang University Science A, 1998, -1(-1): .
@article{title="Machine learning for soil parameter inversion enhanced with Bayesian optimization",
author="Anfeng HU1, Chi WANG1,2, Senlin XIE1, Zhirong XIAO3, Tang LI1, Ang XU1",
journal="Journal of Zhejiang University Science A",
volume="-1",
number="-1",
pages="",
year="1998",
publisher="Zhejiang University Press & Springer",
doi="10.1631/jzus.A2400568"
}
%0 Journal Article
%T Machine learning for soil parameter inversion enhanced with Bayesian optimization
%A Anfeng HU1
%A Chi WANG1
%A 2
%A Senlin XIE1
%A Zhirong XIAO3
%A Tang LI1
%A Ang XU1
%J Journal of Zhejiang University SCIENCE A
%V -1
%N -1
%P
%@ 1673-565X
%D 1998
%I Zhejiang University Press & Springer
%DOI 10.1631/jzus.A2400568
TY - JOUR
T1 - Machine learning for soil parameter inversion enhanced with Bayesian optimization
A1 - Anfeng HU1
A1 - Chi WANG1
A1 - 2
A1 - Senlin XIE1
A1 - Zhirong XIAO3
A1 - Tang LI1
A1 - Ang XU1
J0 - Journal of Zhejiang University Science A
VL - -1
IS - -1
SP -
EP -
%@ 1673-565X
Y1 - 1998
PB - Zhejiang University Press & Springer
ER -
DOI - 10.1631/jzus.A2400568
Abstract: Machine learning has strong potential for soil settlement prediction, but determining hyperparameters for machine learning models is often intricate and laborious. Therefore, we apply bayesian optimization to determine optimal hyperparameter combinations, enhancing the effectiveness of machine learning models for soil parameter inversion. The machine learning models are trained using numerical simulation data generated with the Modified Cam-Clay model in ABAQUS, and their performance is evaluated using ground set-tlement monitoring data from an airport runway. Five optimized machine learning modelsdecision tree, random forest, support vector regression, deep neural network, and one-dimensional convolutional neural networkare compared in terms of their accuracy for soil parameter inversion and settlement prediction. The results indicate that bayesian optimization efficiently utilizes prior knowledge to identify optimal hyperparameters, significantly improving model performance. Among the evaluated models, the one-dimensional convolutional neural network achieves the highest accuracy in soil parameter inversion, generating settlement predictions that closely match real monitoring data. These findings demonstrate the effectiveness of the proposed approach for soil parameter inversion and settlement prediction, and how bayesian optimization can refine the model selection process.
Open peer comments: Debate/Discuss/Question/Opinion
<1>