On the Training Algorithms for Artificial Neural Network in Predicting Compressive Strength of Recycled Aggregate Concrete

Due to the different difficulties of compressive strength of recycled aggregate concrete (RAC) prediction, this investigation develops a prediction architecture based on machine learning algorithms. The artificial neural networks algorithm and artificial neural networkswith a cascade-correlation algorithm using one hidden layer or two hidden layers are proposed to predict the compressive strength of the recycled aggregate concrete. In this research, 112 datasets of recycled aggregate concrete are gathered from the literature with 6 inputs. Moreover, this investigation has predicted the age effect of recycled aggregate age on the compressive strength of recycled aggregate concrete. The reliability of ANN architecture is evaluated by some criteria such as correlation coefficient (R), root mean square error (RMSE) and mean absolute error (MAE). The best ANN architecture could be considered as a new tool for an estimation of the RAC compressive strength.