Prediction of the compressive strength of alkali-activated geopolymeric concrete binders by neuro-fuzzy modeling: a case study

In the present study, compressive strength of different types of alkali-activated binders has been modeled using adaptive neuro-fuzzy interfacial systems (ANFIS). The model was constructed from 395 experimental data collected from the literature and divided into 80% and 20% for training and testing phases, respectively. Absolute fraction of variance, absolute percentage error and root mean square error of both training and testing phases showed relatively high accuracy of the proposed ANFIS model. From the obtained results, a comparative study was performed to show the interaction of some selected factors on the compressive strength of the alkali-activated binders considered. The findings were in accordance with the experimental studies and those results presented in the literature