Prediction of Compressive Strength of Self-Compacting Concrete Containing Fly Ash Using Multivariate Adaptive Regression Splines (MARS)

Document Type : Research Paper

Authors

1 PhD student, Babol Noshirvani University of Technology

2 Department of Civil Engineering, Tabari University of Babol

3 Assistant Professor, Tabari University of Babol

Abstract

Utilization of Self-compacting concrete can reduce expenses of the construction and time, therefore the use of artificial intelligence methods to estimation of concrete properties seems necessery .The main purpose of the study presented in this paper was to investigate the feasibility of using multivariate adaptive regression spline (MARS) for the prediction of 28-day compressive strength of self-compacting concrete with an optimal mixing ratio. Total of 94 dataset collected from the published paper were used in this study. To compare the performance of the technique, prediction was also done using a multilayer perceptron neural network model. MARS model in the training phase model (RMSE=4/250) was better performance than Neural Network (RMSE=4/626). The results of error indices of the testing stage in MARS and Neural Network methods respectively (RMSE=3/007) and (RMSE=4/049) were performed accurately in compressive strength prediction The analysis indicated that the proposed MARS model can gain a high precision, which provides an alternative method for predicting the properties of SCC.

Keywords

Files

Share

How to cite

Statistics