Comparative study of machine learning methods to predict compressive strength of high-performance concrete and model validation on experimental data

Concrete compressive strength (CCS) is one of the most important parameters to determine the performance of concrete during service conditions. To accurately predict the compressive strength of the entire concrete system makes it a great challenge for a sustainable built environment and future generations since the materials are randomly distributed materials throughout the concrete. In this study, a comparative analysis for predicting the compressive strength of high-performance concrete has been carried out using various machine learning methods. Further, the top-performing models are hyper-parameter optimized to improve the accuracy of the model. To understand the importance of each feature in the trained model, feature selection is done based on the best-performing model. The result indicates that the Gradient Boosted Tree algorithm performs best with 0.94 R ², and the most important features for concrete compressive strength prediction are the age of concrete, cement, and water, and the least important feature is coarse aggregate. Hence, the Gradient Boosted Tree algorithm can be used to predict the compressive strength of concrete which helps the contractors to reduce the cost and time in concrete mix designing and prevent the unnecessary wastage of material caused by numerous mixture trials.