Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization
Accurate assessment of undrained shear strength (USS) for soft sensitive clays is a great concern in geotechnical engineering practice. This study applies novel data-driven extreme gradient boosting (XGBoost) and random forest (RF) ensemble learning methods for capturing the relationships between th...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2021-01-01
|
| Series: | Geoscience Frontiers |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1674987120300669 |
| _version_ | 1797713085133225984 |
|---|---|
| author | Wengang Zhang Chongzhi Wu Haiyi Zhong Yongqin Li Lin Wang |
| author_facet | Wengang Zhang Chongzhi Wu Haiyi Zhong Yongqin Li Lin Wang |
| author_sort | Wengang Zhang |
| collection | DOAJ |
| description | Accurate assessment of undrained shear strength (USS) for soft sensitive clays is a great concern in geotechnical engineering practice. This study applies novel data-driven extreme gradient boosting (XGBoost) and random forest (RF) ensemble learning methods for capturing the relationships between the USS and various basic soil parameters. Based on the soil data sets from TC304 database, a general approach is developed to predict the USS of soft clays using the two machine learning methods above, where five feature variables including the preconsolidation stress (PS), vertical effective stress (VES), liquid limit (LL), plastic limit (PL) and natural water content (W) are adopted. To reduce the dependence on the rule of thumb and inefficient brute-force search, the Bayesian optimization method is applied to determine the appropriate model hyper-parameters of both XGBoost and RF. The developed models are comprehensively compared with three comparison machine learning methods and two transformation models with respect to predictive accuracy and robustness under 5-fold cross-validation (CV). It is shown that XGBoost-based and RF-based methods outperform these approaches. Besides, the XGBoost-based model provides feature importance ranks, which makes it a promising tool in the prediction of geotechnical parameters and enhances the interpretability of model. |
| first_indexed | 2024-03-12T07:31:16Z |
| format | Article |
| id | doaj.art-7613fdbaaf514437b1236d3e31c5580c |
| institution | Directory Open Access Journal |
| issn | 1674-9871 |
| language | English |
| last_indexed | 2024-03-12T07:31:16Z |
| publishDate | 2021-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Geoscience Frontiers |
| spelling | doaj.art-7613fdbaaf514437b1236d3e31c5580c2023-09-02T21:46:19ZengElsevierGeoscience Frontiers1674-98712021-01-01121469477Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimizationWengang Zhang0Chongzhi Wu1Haiyi Zhong2Yongqin Li3Lin Wang4School of Civil Engineering, Chongqing University, Chongqing 400045, ChinaSchool of Civil Engineering, Chongqing University, Chongqing 400045, ChinaSchool of Civil Engineering, Chongqing University, Chongqing 400045, ChinaSchool of Civil Engineering, Chongqing University, Chongqing 400045, ChinaCorresponding author.; School of Civil Engineering, Chongqing University, Chongqing 400045, ChinaAccurate assessment of undrained shear strength (USS) for soft sensitive clays is a great concern in geotechnical engineering practice. This study applies novel data-driven extreme gradient boosting (XGBoost) and random forest (RF) ensemble learning methods for capturing the relationships between the USS and various basic soil parameters. Based on the soil data sets from TC304 database, a general approach is developed to predict the USS of soft clays using the two machine learning methods above, where five feature variables including the preconsolidation stress (PS), vertical effective stress (VES), liquid limit (LL), plastic limit (PL) and natural water content (W) are adopted. To reduce the dependence on the rule of thumb and inefficient brute-force search, the Bayesian optimization method is applied to determine the appropriate model hyper-parameters of both XGBoost and RF. The developed models are comprehensively compared with three comparison machine learning methods and two transformation models with respect to predictive accuracy and robustness under 5-fold cross-validation (CV). It is shown that XGBoost-based and RF-based methods outperform these approaches. Besides, the XGBoost-based model provides feature importance ranks, which makes it a promising tool in the prediction of geotechnical parameters and enhances the interpretability of model.http://www.sciencedirect.com/science/article/pii/S1674987120300669Undrained shear strengthExtreme gradient boostingRandom forestBayesian optimizationk-fold CV |
| spellingShingle | Wengang Zhang Chongzhi Wu Haiyi Zhong Yongqin Li Lin Wang Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization Geoscience Frontiers Undrained shear strength Extreme gradient boosting Random forest Bayesian optimization k-fold CV |
| title | Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization |
| title_full | Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization |
| title_fullStr | Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization |
| title_full_unstemmed | Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization |
| title_short | Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization |
| title_sort | prediction of undrained shear strength using extreme gradient boosting and random forest based on bayesian optimization |
| topic | Undrained shear strength Extreme gradient boosting Random forest Bayesian optimization k-fold CV |
| url | http://www.sciencedirect.com/science/article/pii/S1674987120300669 |
| work_keys_str_mv | AT wengangzhang predictionofundrainedshearstrengthusingextremegradientboostingandrandomforestbasedonbayesianoptimization AT chongzhiwu predictionofundrainedshearstrengthusingextremegradientboostingandrandomforestbasedonbayesianoptimization AT haiyizhong predictionofundrainedshearstrengthusingextremegradientboostingandrandomforestbasedonbayesianoptimization AT yongqinli predictionofundrainedshearstrengthusingextremegradientboostingandrandomforestbasedonbayesianoptimization AT linwang predictionofundrainedshearstrengthusingextremegradientboostingandrandomforestbasedonbayesianoptimization |