COMPRESSIVE STRENGTH MODELING OF SCC USING LINEAR REGRESSION AND ARTIFICIAL NEURAL NETWORK APPROACH

The paper presents the comparative performance of the models developed to predict 28-day compressive strength using linear regression approach with artificial neural network approach. The data used in the models was obtained experimentally with various fly ash contents in total powder content of 550 kg/m(3) and bottom ash contents as replacement of fine aggregates in SCC mixes and are arranged in the format of eight input parameters that cover the contents of cement, fine aggregates, fly ash as replacement of cement, bottom ash as replacement of sand, water and water-to-binder ratio with coarse aggregate content kept constant throughout the study and an output parameter which is 28-days compressive strength of concrete. The expression for 28-day compressive strength was developed using linear regression kernels and the performance of the models was compared with that of the artificial neural network to predict the 28-day compressive strength.