Abstract & Overview
Future developments in materials science engineering will be greatly influenced by the application of machine learning for determining the properties of concrete, especially its compressive strength. This research predicts the compressive strength of concrete with eight independent variables, including cement, blast furnace slag, fly ash, water, superplasticizer, coarse aggregate, fine aggregate, and age using supervised machine learning (ML) techniques of linear regression (LR) and light gradient boosting machine (LGBM). The ML models are fed a total of 1030 datasets using a 70:30 split ratio for training and testing. Performance metrics like R2 , MAE, MSE, and RMSE are used to assess how well the ML models are in making predictions. From the research, the LR model (R2 value of 0.607) is less effective than the LGBM model (R2 value of 0.920) in predicting compressive strength. Furthermore, feature importance predicted by LGBM shows that the cement content (2331), fine aggregate (2200), and coarse aggregate (2076) all significantly influence the prediction of concrete compressive strength