Comparative study of soft computing techniques for the prediction of concrete strength containing waste material

Ahmad, Ayaz, Finnegan, William, Jiang, Yadong, & Goggins, Jamie. (2022). Comparative study of soft computing techniques for the prediction of concrete strength containing waste material. Paper presented at the Civil Engineering Research in Ireland (CERI 2022), Dublin, 25-26 August.

Abstract

The application of artificial intelligence algorithms to anticipate the strength properties of various types of concrete is increasing in prominence. This study describes the use of two artificial intelligence algorithms, such as decision tree (DT) and bagging algorithm (BA), to anticipate the compressive strength of concrete containing fly ash. Python instructions were executed on the appropriate models using the anaconda navigator software. The models were conducted with seven input variables (cement, water, fly ash, superplasticizers, coarse aggregate, fine aggregate, and age) and one output parameter (i.e. compressive strength). Results show that the precision level of the BA towards the prediction of concrete’s strength is high compared to the DT model. The said accuracy is indicated by the coefficient of determination value, which equals 0.93 for the BA and 0.86 for the DT model. The statistical checks also verified the accuracy level of the employed algorithms. The low values of the mean absolute error and root mean square error also confirm high accuracy for BA compared to DT.

Funder

Science Foundation Ireland
MaREI Research Centre for Energy, Climate and Marine
Sustainable Energy Authority of Ireland

Publisher

Civil Engineering Research Association of Ireland

Rights

Attribution-NonCommercial-NoDerivatives 4.0 International

cbn

Collections

Civil Engineering (Conference Papers)