To address the issue of strength prediction in concrete in practical engineering,a Stacking ensemble algorithm is adopted,and the water-cement ratio is introduced as a new feature.By combining the density-based clustering algorithm(DBSCAN)and the akaike information criterion(AIC)for base model selection,the DBSCAN-AIC-Stacking(DASA)model is proposed to improve the accuracy of compressive strength prediction in concrete.The prediction results of the DASA model fall within a range of±20％,with the majority of results falling within±10％.This model demonstrates significant advantages in practical engineering and provides a feasible and effective solution to the problem of concrete strength prediction.
维普
万方数据