Predicting Optimum Moisture Content by the individual and hybrid approach of machine learning
The Optimal Moisture Content (OMC) is the moisture level at which soil reaches its maximum density during compaction, making it vital in geotechnical engineering and construction for establishing the best conditions for soil strength and stability in infrastructure. Existing methods to determine OMC...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Tamkang University Press
2025-01-01
|
| Series: | Journal of Applied Science and Engineering |
| Subjects: | |
| Online Access: | http://jase.tku.edu.tw/articles/jase-202508-28-08-0004 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841556043021156352 |
|---|---|
| author | Yinghui Yang Yahui Dai Qunting Yang |
| author_facet | Yinghui Yang Yahui Dai Qunting Yang |
| author_sort | Yinghui Yang |
| collection | DOAJ |
| description | The Optimal Moisture Content (OMC) is the moisture level at which soil reaches its maximum density during compaction, making it vital in geotechnical engineering and construction for establishing the best conditions for soil strength and stability in infrastructure. Existing methods to determine OMC are both expensive and time-intensive. Machine learning offers a promising alternative by enabling the creation of advanced predictive models and algorithms that can improve the accuracy and efficiency of OMC predictions compared to traditional empirical methods. This study aims to develop an advanced framework utilizing the Gaussian Process Regression (GPR) model for predicting OMC. The research demonstrates a strong relationship using GPR models between OMC and important soil parameters such as particle size distribution, linear shrinkage, and the kind and quantity of stabilizing chemicals. To further enhance the predictive accuracy of these models, two meta-heuristic optimization techniques—Atom Search Optimization (ASO) and Reptile Search Algorithm (RSA)—are employed. The study presents three distinct models—GPAS, GPRS, and a standalone GPR model—each providing critical insights for accurate OMC prediction. Among these, the GPAS model demonstrates superior performance, achieving an outstanding R² score of 0.992 and a notably low RMSE of 0.719. These results underscore the GPAS model’s exceptional reliability, precision, and its robust capability in forecasting the outcomes of soil stabilization efforts. Efficient OMC prediction helps in reducing the overuse of stabilizing additives and water, leading to more sustainable construction practices. It also minimizes the environmental impact associated with soil excavation and modification. |
| format | Article |
| id | doaj-art-8275746a725e4e8ba98472a9e610192b |
| institution | Kabale University |
| issn | 2708-9967 2708-9975 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Tamkang University Press |
| record_format | Article |
| series | Journal of Applied Science and Engineering |
| spelling | doaj-art-8275746a725e4e8ba98472a9e610192b2025-01-07T14:36:57ZengTamkang University PressJournal of Applied Science and Engineering2708-99672708-99752025-01-01288 1657166910.6180/jase.202508_28(8).0004Predicting Optimum Moisture Content by the individual and hybrid approach of machine learningYinghui Yang0Yahui Dai1Qunting Yang2College of Information and Engineering, Henan University of Animal Husbandry and Economy; Zhengzhou Henan, 450045, ChinaScience and Technology Research Institute of State Power Investment Group Co., LTD. Beijing,102209, ChinaCollege of Air Traffic Management, Civil Aviation University of China; Tianjin, 300300, ChinaThe Optimal Moisture Content (OMC) is the moisture level at which soil reaches its maximum density during compaction, making it vital in geotechnical engineering and construction for establishing the best conditions for soil strength and stability in infrastructure. Existing methods to determine OMC are both expensive and time-intensive. Machine learning offers a promising alternative by enabling the creation of advanced predictive models and algorithms that can improve the accuracy and efficiency of OMC predictions compared to traditional empirical methods. This study aims to develop an advanced framework utilizing the Gaussian Process Regression (GPR) model for predicting OMC. The research demonstrates a strong relationship using GPR models between OMC and important soil parameters such as particle size distribution, linear shrinkage, and the kind and quantity of stabilizing chemicals. To further enhance the predictive accuracy of these models, two meta-heuristic optimization techniques—Atom Search Optimization (ASO) and Reptile Search Algorithm (RSA)—are employed. The study presents three distinct models—GPAS, GPRS, and a standalone GPR model—each providing critical insights for accurate OMC prediction. Among these, the GPAS model demonstrates superior performance, achieving an outstanding R² score of 0.992 and a notably low RMSE of 0.719. These results underscore the GPAS model’s exceptional reliability, precision, and its robust capability in forecasting the outcomes of soil stabilization efforts. Efficient OMC prediction helps in reducing the overuse of stabilizing additives and water, leading to more sustainable construction practices. It also minimizes the environmental impact associated with soil excavation and modification.http://jase.tku.edu.tw/articles/jase-202508-28-08-0004machine learninggaussian process regressionmetaheuristic algorithmsoptimum moisture contentartificial intelligence |
| spellingShingle | Yinghui Yang Yahui Dai Qunting Yang Predicting Optimum Moisture Content by the individual and hybrid approach of machine learning Journal of Applied Science and Engineering machine learning gaussian process regression metaheuristic algorithms optimum moisture content artificial intelligence |
| title | Predicting Optimum Moisture Content by the individual and hybrid approach of machine learning |
| title_full | Predicting Optimum Moisture Content by the individual and hybrid approach of machine learning |
| title_fullStr | Predicting Optimum Moisture Content by the individual and hybrid approach of machine learning |
| title_full_unstemmed | Predicting Optimum Moisture Content by the individual and hybrid approach of machine learning |
| title_short | Predicting Optimum Moisture Content by the individual and hybrid approach of machine learning |
| title_sort | predicting optimum moisture content by the individual and hybrid approach of machine learning |
| topic | machine learning gaussian process regression metaheuristic algorithms optimum moisture content artificial intelligence |
| url | http://jase.tku.edu.tw/articles/jase-202508-28-08-0004 |
| work_keys_str_mv | AT yinghuiyang predictingoptimummoisturecontentbytheindividualandhybridapproachofmachinelearning AT yahuidai predictingoptimummoisturecontentbytheindividualandhybridapproachofmachinelearning AT quntingyang predictingoptimummoisturecontentbytheindividualandhybridapproachofmachinelearning |