Prediction of shear strength in UHPC beams using machine learning-based models and SHAP interpretation

Construction and Building Materials, Journal Year: 2023, Volume and Issue: 408, P. 133752 - 133752

Published: Oct. 14, 2023

Language: Английский

---

Predictive modeling for compressive strength of 3D printed fiber-reinforced concrete using machine learning algorithms

Case Studies in Construction Materials, Journal Year: 2023, Volume and Issue: 20, P. e02728 - e02728

Published: Nov. 30, 2023

Three-dimensional (3D) printing in the construction industry is growing rapidly due to its inherent advantages, including intricate geometries, reduced waste, accelerated construction, cost-effectiveness, eco-friendliness, and improved safety. However, optimizing mixture composition for 3D-printed concrete remains a formidable task, encompassing multiple variables requiring comprehensive trial-and-error experimentation process. Accordingly, this study used seven machine learning (ML) algorithms, support vector regression (SVR), decision tree (DT), SVR-Bagging, SVR-Boosting, random forest (RF), gradient boosting (GB), gene expression programming (GEP) forecasting compressive strength (CS) of 3D printed fiber-reinforced (3DP-FRC). For model development, 299 data points were collected from experimental studies split into two portions: 70% training 30% validation. Various statistical metrics employed examine accuracy generalizability established models. The DT, RF, GB, GEP models demonstrated higher validation set, achieving correlation (R) values 0.987, 0.986, 0.98, respectively. exhibited mean absolute error (MAE) scores 4.644, 3.989, 3.90, 5.691, Furthermore, combination SVR with bagging techniques slightly compared individual model. Additionally, Shapley Additive exPlanations (SHAP) approach unveils proportional significance parameters influencing CS 3DP-FRC. SHAP technique revealed that water, silica fume, superplasticizer, sand content, loading directions are dominant estimating local interpretability intrinsic relationship between diverse input their impacts on offers significant insights optimum mix proportion

Language: Английский

---

Compressive strength prediction of sustainable concrete incorporating rice husk ash (RHA) using hybrid machine learning algorithms and parametric analyses

Case Studies in Construction Materials, Journal Year: 2024, Volume and Issue: 20, P. e03030 - e03030

Published: March 5, 2024

The construction industry is making efforts to reduce the environmental impact of cement production in concrete by incorporating alternative and supplementary cementitious materials, as well lowering carbon emissions. One such material that has gained popularity this context rice husk ash (RHA) due its pozzolanic reactions. This study aims forecast compressive strength (CS) RHA-based (RBC) examining effects several factors cement, RHA content, curing age, water usage, aggregate amount, superplasticizer content. To accomplish this, collected analyzed data from literature, resulting a dataset 1404 observations. Several machine learning (ML) models, light gradient boosting (LGB), extreme (XGB), random forest (RF), hybrid (HML) approaches like XGB-LGB XGB-RF were employed thoroughly analyze these parameters assess their on strength. was split into training testing groups, statistical analyses performed determine relationships between input CS. Moreover, performance all models evaluated using various evaluation criteria, including mean absolute percentage error (MAPE), coefficient efficiency (CE), root square (RMSE), determination (R2). model found have higher precision (R2 = 0.95, RMSE 5.255 MPa) compared other models. SHAP (SHapley Additive exPlanations) analysis revealed RHA, had positive effect Overall, study's findings suggest with identified can be used accurately predict CS RBC. application technologies sector facilitate rapid low-cost identification qualities parameters.

Language: Английский

---

Estimating compressive strength of concrete containing rice husk ash using interpretable machine learning-based models

Case Studies in Construction Materials, Journal Year: 2024, Volume and Issue: 20, P. e02901 - e02901

Published: Jan. 19, 2024

The construction sector is a major contributor to global greenhouse gas emissions. Using recycled and waste materials in concrete practical solution address environmental challenges. Currently, agricultural widely used as substitute for cement the production of eco-friendly concrete. However, traditional methods assessing strength such are both expensive time-consuming. Therefore, this study uses machine learning techniques develop prediction models compressive (CS) rice husk ash (RHA) ML present include random forest (RF), light gradient boosting (LightGBM), ridge regression, extreme (XGBoost). A total 348 values CS were collected from experimental studies, five characteristics RHA taken input variables. For performance assessment models, multiple statistical metrics used. During training phase, correlation coefficients (R) obtained RF, XGBoost, LightGBM 0.943, 0.981, 0.985, 0.996, respectively. In testing set, these demonstrated even higher performance, with 0.971, 0.993, 0.992, 0.998 LightGBM, analysis revealed that model outperformed other whereas regression exhibited comparatively lower accuracy. SHapley Additive exPlanation (SHAP) method was employed interpretability developed model. SHAP water-to-cement controlling parameter estimating conclusion, provides valuable guidance builders researchers estimate it suggested more variables be incorporated hybrid utilized further enhance reliability precision models.

Language: Английский

---

Predictive models in machine learning for strength and life cycle assessment of concrete structures

Automation in Construction, Journal Year: 2024, Volume and Issue: 162, P. 105412 - 105412

Published: April 3, 2024

Language: Английский

---

Predicting Penetration Depth in Ultra-High-Performance Concrete Targets under Ballistic Impact: An Interpretable Machine Learning Approach Augmented by Deep Generative Adversarial Network

Results in Engineering, Journal Year: 2025, Volume and Issue: unknown, P. 103909 - 103909

Published: Jan. 1, 2025

Language: Английский

---

Compressive strength evaluation of cement-based materials in sulphate environment using optimized deep learning technology

Developments in the Built Environment, Journal Year: 2023, Volume and Issue: 16, P. 100298 - 100298

Published: Dec. 1, 2023

Strength serves as a vital performance metric for assessing long-term durability of cement-based materials. Nevertheless, there is scarcity models available predicting residual strength in-situ structures made materials exposed to sulphate conditions. To address this challenge, study presents novel approach using deep learning predict the degradation compressive under marine environments. Specifically, convolutional neural network (DCNN) established, consisting two layers, one pooling layer, and fully connected layers. In innovative model, contents cement, water-to-cement ratio, sand, concentration exposure temperature are selected inputs, while output subjected deterioration. improve forecast capability, particle swarm optimization adopted optimizing hyperparameters DCNN, which can be implemented by reducing discrepancy between model prediction measured strength. Finally, experimental data used establish evaluate proposed method. The results show that learning-based predictive has best suffering from attack via comparison with other commonly models. outcome research offers potential solution remaining undergo practical attack.

Language: Английский

---

Application of metaheuristic optimization algorithms in predicting the compressive strength of 3D-printed fiber-reinforced concrete

Developments in the Built Environment, Journal Year: 2023, Volume and Issue: 17, P. 100307 - 100307

Published: Dec. 22, 2023

In recent years, the construction industry has been striving to make production faster and handle more complex architectural designs. Waste reduction, geometric freedom, lower costs, speedy 3D-printed fiber-reinforced concrete (3DPFRC) alternative for future construction. However, achieving optimum mixture composition 3DPFRC remains a daunting task, entailing consideration of multiple variables necessitating an extensive trial-and-error experimental process. Therefore, this study investigated application different metaheuristic optimization algorithms predict compressive strength (CS) 3DPFRC. A database 299 data samples with 16 input features was compiled from studies in literature. Six algorithms, such as human felicity algorithm (HFA), differential evolution (DEA), nuclear reaction (NRO), Harris hawks (HHO), lightning search (LSA), tunicate swarm (TSA) were applied identify optimal hyperparameter combination random forest (RF) model predicting CS Different statistical metrics 10-fold cross-validation used evaluate accuracy models. The TSA-RF exhibited superior performance compared other models, correlation (R), mean absolute error (MAE), root square (RMSE) values 0.99, 2.10 MPa, 3.59 respectively. LSA-RF also performed well, R, MAE, RMSE 2.93 6.23 SHapley Additive exPlanation (SHAP) interpretability elucidates intricate relationships between their effects on CS, thereby offering invaluable insights performance-based mix proportion design

Language: Английский

---

Data-driven Models for Predicting Compressive Strength of 3D-printed Fiber-Reinforced Concrete using Interpretable Machine Learning Algorithms

Case Studies in Construction Materials, Journal Year: 2024, Volume and Issue: 21, P. e03935 - e03935

Published: Nov. 1, 2024

Language: Английский

---

Quantifying compressive strength in limestone powder incorporated concrete with incorporating various machine learning algorithms with SHAP analysis

Asian Journal of Civil Engineering, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 14, 2024

Language: Английский

---