A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Predicting Dry Shrinkage Using Machine Learning Methods
Modeling drying shrinkage presents significant challenges due to the complexity and multitude of contributing parameters. This study provides detailed insights into the input requirements and predictive capabilities of established models by leveraging various datasets from the NU-ITI database. Initially, the performance of a shrinkage model was evaluated. The data for a machine learning random forest model included eight variables, interpreted through SHapley Additive exPlanations (SHAP), which elucidates the most influential inputs. However, the partial dependency graphs yielded minimal information on their relative impacts. This research demonstrates that enhancements in the random forest model’s predictive accuracy improved shrinkage predictions by 25%. This advancement significantly mitigates potential deviations in anticipated strains and stresses. The findings from this comprehensive analysis facilitate the selection and prediction of drying shrinkage, focusing on the most critical factors to ensure the highest accuracy.
Predicting Dry Shrinkage Using Machine Learning Methods
Modeling drying shrinkage presents significant challenges due to the complexity and multitude of contributing parameters. This study provides detailed insights into the input requirements and predictive capabilities of established models by leveraging various datasets from the NU-ITI database. Initially, the performance of a shrinkage model was evaluated. The data for a machine learning random forest model included eight variables, interpreted through SHapley Additive exPlanations (SHAP), which elucidates the most influential inputs. However, the partial dependency graphs yielded minimal information on their relative impacts. This research demonstrates that enhancements in the random forest model’s predictive accuracy improved shrinkage predictions by 25%. This advancement significantly mitigates potential deviations in anticipated strains and stresses. The findings from this comprehensive analysis facilitate the selection and prediction of drying shrinkage, focusing on the most critical factors to ensure the highest accuracy.
Predicting Dry Shrinkage Using Machine Learning Methods
Lecture Notes in Civil Engineering
Kioumarsi, Mahdi (editor) / Shafei, Behrouz (editor) / Khodabandeh, Peyman (author) / Azarhomayun, Fazel (author) / Shekarchi, Mohammad (author) / Kioumarsi, Mahdi (author)
The International Conference on Net-Zero Civil Infrastructures: Innovations in Materials, Structures, and Management Practices (NTZR) ; 2024 ; Oslo, Norway
The 1st International Conference on Net-Zero Built Environment ; Chapter: 61 ; 731-738
2025-01-09
8 pages
Article/Chapter (Book)
Electronic Resource
English
Rule for predicting concrete shrinkage
Engineering Index Backfile | 1931
Predicting submerged hydraulic jump characteristics using machine learning methods
| DOAJ | 2021
Machine Learning Methods for Predicting Soil Compression Index
| Springer Verlag | 2024