Machine learning-based approach for predicting the consolidation characteristics of soft soil

Singh, Moirangthem Johnson1; Kaushik, Anshul1; Patnaik, Gyanesh1; Xu, Dong-Sheng2; Feng, Wei-Qiang3; Rajput, Abhishek1; Prakash, Guru1; Borana, Lalit1

2023-03-01

10.1080/1064119X.2023.2193174

MARINE GEORESOURCES & GEOTECHNOLOGY   影响因子和分区

1064-119X

1521-0618

In recent times, large-scale infrastructural projects are being constructed on varieties of soil, especially in highly compressible marine clays and soft soil. The coefficient of consolidation (c(v)) is one of the most important technical parameters used to estimate the consolidation characteristics of the soil. The experimental laboratory techniques used to obtain c(v) are time-consuming and possess different practical limitations. In this study, a reliable method for predicting c(v) is presented based on machine learning (ML). The study considered 11 inherent soil variables, among which the least significant variables are discarded using univariate feature selection technique. Different ML models were developed like the random forest, artificial neural network, and support vector machine for nonlinear mapping of the c(v) using dimensionally reduced independent variables. Verification against experimental data demonstrates that the Random Forest model accurately predicts the c(v) (with MAE = 0.0231, MSE= 0.00148, and RMSE = 0.03854). Further, a comparative study of the proposed model is presented with available empirical equations and numerically simulated data. Moreover, the strengths and shortcomings of different ML algorithms are also discussed in detail.

Marine clays and soft soil consolidation coefficient of consolidation machine learning

SCI ; EI

英语

其他

Engineering ; Oceanography ; Mining & Mineral Processing

Engineering, Ocean ; Engineering, Geological ; Oceanography ; Mining & Mineral Processing

WOS:000971546700001

TAYLOR & FRANCIS INC

20231714009908

Forestry ; Neural networks ; Support vector machines

Soils and Soil Mechanics:483.1 ; Computer Software, Data Handling and Applications:723 ; Agricultural Equipment and Methods; Vegetation and Pest Control:821

GEOSCIENCES

Web of Science

1.Indian Inst Technol Indore, Civil Engn Dept, Indore 452020, Madhya Pradesh, India
2.Wuhan Univ Technol, Sch Civil Engn & Architecture, Wuhan, Hubei, Peoples R China
3.Southern Univ Sci & Technol, Dept Ocean Sci & Engn, Shenzhen, Guangdong, Peoples R China

Singh, Moirangthem Johnson,Kaushik, Anshul,Patnaik, Gyanesh,et al. Machine learning-based approach for predicting the consolidation characteristics of soft soil[J]. MARINE GEORESOURCES & GEOTECHNOLOGY,2023.

Singh, Moirangthem Johnson.,Kaushik, Anshul.,Patnaik, Gyanesh.,Xu, Dong-Sheng.,Feng, Wei-Qiang.,...&Borana, Lalit.(2023).Machine learning-based approach for predicting the consolidation characteristics of soft soil.MARINE GEORESOURCES & GEOTECHNOLOGY.

Singh, Moirangthem Johnson,et al."Machine learning-based approach for predicting the consolidation characteristics of soft soil".MARINE GEORESOURCES & GEOTECHNOLOGY (2023).