Prediction of SFRSCC's T50 Performance Using Regression Tree Algorithm

Abstract

Concrete is an important building material which has been in human life for more than a century and is used continuously. Despite its age, many properties are still not known. There are different types of concrete according to the materials in the mixture. One of these types of concrete, called special concrete, is steel fiber reinforced self-compacting concrete (SFRSCC). There are different criteria for measuring the performance of concrete. Since the properties are not known exactly, different experiments are carried out in order to provide the characteristics of the requested concrete. In this study, the model was designed by using regression tree method to predict the performance of T50 of SFRSCC. In the designed model using regression tree algorithm, R-squared value was calculated as 0.7341, root mean square error (RMSE) value was calculated as 1.2636, mean absolute error (MAE) value was calculated as 0.8617. © 2019 IEEE.