Compressive Strength Prediction Using Coupled Deep Learning Model with Extreme Gradient Boosting Algorithm: Environmentally Friendly Concrete Incorporating Recycled Aggregate

The application of recycled aggregate as a sustainable material in construction projects is considered promising approach to decrease the carbon footprint concrete structures. Prediction compressive strength (CS) environmentally friendly (EF) containing important for understanding structures’ behaviour. In this research, capability deep learning neural network (DLNN) examined on simulation CS EF concrete. developed compared well-known artificial intelligence (AI) approaches named multivariate adaptive regression spline (MARS), extreme machines (ELMs), and random forests (RFs). dataset was divided into three scenarios 70%-30%, 80%-20%, 90%-10% training/testing explore impact data division percentage capacity AI model. Extreme gradient boosting (XGBoost) integrated with models select influencing variables prediction. Several statistical measures graphical methods were generated evaluate efficiency presented models. regard, results confirmed that DLNN model attained highest value prediction performance minimal root mean squared error (RMSE = 2.23). study revealed could be by increasing number problem using division. demonstrated robustness over other handling complex behaviour Due high accuracy model, method can used practical future use