Prediction of Ecofriendly Concrete Compressive Strength Using Gradient Boosting Regression Tree Combined with...

Area of Science:

• Materials Science
• Civil Engineering
• Computational Science

Background:

• The compressive strength (Cs) of eco-friendly concrete is a key parameter for sustainable building design.
• Accurate prediction of Cs is essential for optimizing concrete mixtures and ensuring structural integrity.
• Existing models may lack the precision required for complex eco-friendly concrete formulations.

Purpose of the Study:

• To develop and validate a highly accurate predictive model for the compressive strength of eco-friendly concrete.
• To enhance prediction precision by employing a hybrid Gradient Boosting Regression Tree (GBRT) model optimized with Grid Search Cross-Validation (GridSearchCV).
• To identify and quantify the influence of key input factors on concrete compressive strength.

Main Methods:

• A dataset comprising 164 experiments on eco-friendly concrete was compiled from previous research.
• A hybrid Gradient Boosting Regression Tree (GBRT) model was implemented and optimized using GridSearchCV.
• Model performance was evaluated using Root Mean Square Error (RMSE) and Coefficient of Determination (R²).
• The Shapley Additive Explanation (SHAP) approach was utilized to interpret model predictions and feature importance.

Main Results:

• The GridSearchCV optimization significantly improved the hyperparameter tuning for the GBRT model compared to default settings.
• The optimized GSC-GBRT model demonstrated robust generalization capabilities.
• The predictive model achieved a Root Mean Square Error (RMSE) of 2.3214 and a Coefficient of Determination (R²) of 0.9612 on the testing dataset.
• SHAP analysis provided insights into the significance and contribution of input variables (W/B ratio, age, RA%, GGBFS%, superplasticizer) to compressive strength.

Conclusions:

• The proposed GSC-GBRT model offers a reliable and advantageous approach for predicting the compressive strength of eco-friendly concrete.
• The optimized model enhances prediction accuracy, supporting the efficient design of sustainable buildings.
• Understanding the influence of input factors through SHAP analysis aids in material selection and mix design for improved concrete performance.