Compression Index Regression of Fine-Grained Soils with Machine Learning Algorithms

Soil consolidation, particularly in fine-grained soils like clay, is crucial in predicting settlement and ensuring the stability of structures. Additionally, the compressibility of fine-grained soils is of critical importance not only in civil engineering but also in various other fields of study. T...

Full description

Saved in:
Bibliographic Details
Main Authors: Mintae Kim, Muharrem A. Senturk, Liang Li
Format: Article
Language:English
Published: MDPI AG 2024-09-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/14/19/8695
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850283885841088512
author Mintae Kim
Muharrem A. Senturk
Liang Li
author_facet Mintae Kim
Muharrem A. Senturk
Liang Li
author_sort Mintae Kim
collection DOAJ
description Soil consolidation, particularly in fine-grained soils like clay, is crucial in predicting settlement and ensuring the stability of structures. Additionally, the compressibility of fine-grained soils is of critical importance not only in civil engineering but also in various other fields of study. The compression index (<i>C<sub>c</sub></i>), derived from soil properties such as the liquid limit (LL), plastic limit (PL), plasticity index (PI), water content (<i>w</i>), initial void ratio (<i>e</i><sub>0</sub>), and specific gravity (<i>G<sub>s</sub></i>), plays a vital role in understanding soil behavior. This study employs machine learning algorithms—the random forest regressor (RFR), gradient boosting regressor (GBR), and AdaBoost regressor (ABR)—to predict the <i>C<sub>c</sub></i> values based on a dataset comprising 915 samples. The dataset includes LL, PL, <i>W</i>, PI, <i>G<sub>s</sub></i>, and <i>e</i><sub>0</sub> as the inputs, with <i>C<sub>c</sub></i> as the output parameter. The algorithms are trained and evaluated using metrics such as the coefficient of determination (R<sup>2</sup>), mean absolute error (MAE), mean squared error (MSE), and root mean squared error (RMSE). Hyperparameter optimization is performed to enhance the model performance. The best-performing model, the GBR model, achieves a training R<sup>2</sup> of 0.925 and a testing R<sup>2</sup> of 0.930 with the input combination [<i>w</i>, PL, LL, PI, <i>e</i><sub>0</sub>, <i>G<sub>s</sub></i>]. The RFR model follows closely, with a training R<sup>2</sup> of 0.970 and a testing R<sup>2</sup> of 0.926 using the same input combination. The ABR model records a training R<sup>2</sup> of 0.847 and a testing R<sup>2</sup> of 0.921 under similar conditions. These results indicate superior predictive accuracy compared to previous studies using traditional statistical and machine learning methods. Machine learning algorithms, specifically the gradient boosting regressor and random forest regressor, demonstrate substantial potential in predicting the <i>C<sub>c</sub></i> value for fine-grained soils based on multiple soil parameters. This study involves leveraging the efficiency and effectiveness of these algorithms in geotechnical engineering applications, offering a promising alternative to traditional oedometer testing methods. Accurately predicting the compression index can significantly aid in the assessment of soil settlement and the design of stable foundations, thereby reducing the time and costs associated with laboratory testing.
format Article
id doaj-art-0eee379d76cf49bf824788bfa8d0e446
institution OA Journals
issn 2076-3417
language English
publishDate 2024-09-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-0eee379d76cf49bf824788bfa8d0e4462025-08-20T01:47:41ZengMDPI AGApplied Sciences2076-34172024-09-011419869510.3390/app14198695Compression Index Regression of Fine-Grained Soils with Machine Learning AlgorithmsMintae Kim0Muharrem A. Senturk1Liang Li2School of Civil, Environmental, and Architectural Engineering, Korea University, Seoul 02841, Republic of KoreaDepartment of Computer Engineering, Istanbul University, Istanbul 34452, TurkeyGeotechnical Department, Terracon Consultants Inc., Midland, TX 79707, USASoil consolidation, particularly in fine-grained soils like clay, is crucial in predicting settlement and ensuring the stability of structures. Additionally, the compressibility of fine-grained soils is of critical importance not only in civil engineering but also in various other fields of study. The compression index (<i>C<sub>c</sub></i>), derived from soil properties such as the liquid limit (LL), plastic limit (PL), plasticity index (PI), water content (<i>w</i>), initial void ratio (<i>e</i><sub>0</sub>), and specific gravity (<i>G<sub>s</sub></i>), plays a vital role in understanding soil behavior. This study employs machine learning algorithms—the random forest regressor (RFR), gradient boosting regressor (GBR), and AdaBoost regressor (ABR)—to predict the <i>C<sub>c</sub></i> values based on a dataset comprising 915 samples. The dataset includes LL, PL, <i>W</i>, PI, <i>G<sub>s</sub></i>, and <i>e</i><sub>0</sub> as the inputs, with <i>C<sub>c</sub></i> as the output parameter. The algorithms are trained and evaluated using metrics such as the coefficient of determination (R<sup>2</sup>), mean absolute error (MAE), mean squared error (MSE), and root mean squared error (RMSE). Hyperparameter optimization is performed to enhance the model performance. The best-performing model, the GBR model, achieves a training R<sup>2</sup> of 0.925 and a testing R<sup>2</sup> of 0.930 with the input combination [<i>w</i>, PL, LL, PI, <i>e</i><sub>0</sub>, <i>G<sub>s</sub></i>]. The RFR model follows closely, with a training R<sup>2</sup> of 0.970 and a testing R<sup>2</sup> of 0.926 using the same input combination. The ABR model records a training R<sup>2</sup> of 0.847 and a testing R<sup>2</sup> of 0.921 under similar conditions. These results indicate superior predictive accuracy compared to previous studies using traditional statistical and machine learning methods. Machine learning algorithms, specifically the gradient boosting regressor and random forest regressor, demonstrate substantial potential in predicting the <i>C<sub>c</sub></i> value for fine-grained soils based on multiple soil parameters. This study involves leveraging the efficiency and effectiveness of these algorithms in geotechnical engineering applications, offering a promising alternative to traditional oedometer testing methods. Accurately predicting the compression index can significantly aid in the assessment of soil settlement and the design of stable foundations, thereby reducing the time and costs associated with laboratory testing.https://www.mdpi.com/2076-3417/14/19/8695compression indexmachine learningrandom forestgradient boostingAdaBoost
spellingShingle Mintae Kim
Muharrem A. Senturk
Liang Li
Compression Index Regression of Fine-Grained Soils with Machine Learning Algorithms
Applied Sciences
compression index
machine learning
random forest
gradient boosting
AdaBoost
title Compression Index Regression of Fine-Grained Soils with Machine Learning Algorithms
title_full Compression Index Regression of Fine-Grained Soils with Machine Learning Algorithms
title_fullStr Compression Index Regression of Fine-Grained Soils with Machine Learning Algorithms
title_full_unstemmed Compression Index Regression of Fine-Grained Soils with Machine Learning Algorithms
title_short Compression Index Regression of Fine-Grained Soils with Machine Learning Algorithms
title_sort compression index regression of fine grained soils with machine learning algorithms
topic compression index
machine learning
random forest
gradient boosting
AdaBoost
url https://www.mdpi.com/2076-3417/14/19/8695
work_keys_str_mv AT mintaekim compressionindexregressionoffinegrainedsoilswithmachinelearningalgorithms
AT muharremasenturk compressionindexregressionoffinegrainedsoilswithmachinelearningalgorithms
AT liangli compressionindexregressionoffinegrainedsoilswithmachinelearningalgorithms