Optimizing Sustainable Construction Materials with Machine Learning Algorithms: Predicting Compressive Strength of Concrete Composites

This research paper presents a study on predicting the compressive strength (CS) of Limestone and Calcined Clay-based concrete composites using machine learning algorithms. Limestone and Calcined Clay are promising materials to replace Ordinary Portland Cement, with the potential benefits of significantly reduced carbon dioxide emissions and lower production costs. In this study, three Ensemble Machine Learning (EML) models, Gradient Boosting, Random Forest, and AdaBoost, were employed to develop predictive models for the compressive strength of the concrete composite. The models were trained using 80% of the data and tested with the remaining data. The results showed that the developed models effectively predicted the compressive strength of concrete composite with high accuracy and consistency. The findings of this research can provide valuable insights into the development of sustainable construction materials and the use of machine learning techniques in predicting the strength of concrete composites. The assessment of model efficiency revealed that the Gradient Boosting model emerged as the optimal choice for achieving accurate CS predictions, demonstrating a superior Correlation Coefficient (R²) alongside diminished values of Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). The Random Forest model was deemed inferior with lower R² and higher RMSE and MAE values.