Development of multiple linear regression, artificial neural networks and fuzzy logic models to predict the efficiency factor and durability indicator of nano natural pozzolana as cement additive

Abstract

The current study aims at predicting the efficiency factor (EF) and durability indicator (DI) of the natural pozzolana when added as a cement replacement at nano scale. Multiple linear regression (MLR), artificial neural networks (ANN) and fuzzy logic (FL) tools were employed in the analytical study. The studied sample was the data collected from an experimental study carried out by the authors, on a local natural pozzolana. Curing time, nano natural pozzoalna (NNP) content, median particle size of natural pozzolana, water/binder (w/b) ratio and the superplasticizer dosage were selected as input variables. The curing times were investigated at 2, 7, 28, 90 and 180 days. NNP was added at six percentages, i.e. 0, 1, 2, 3, 4 and 5%. Two median particle sizes i.e. 100 and 500 nm, were studied. Four w/b ratios and four superplasticizer dosages were examined, i.e. 0.4, 0.5, 0.6 and 0.7, and 0, 0.5, 2 and 4 l/m3, respectively. The correlation coefficients and several performance criteria were computed to evaluate the developed models. Based on the analysis results, it can be concluded that EF and DI of NNP can be effectively predicted using ANN and FL techniques. The results obtained by MLR were far from those obtained by both ANN and FL. In addition, ANN tool was slightly more accurate than FL as far as prediction of EF is concerned. Coefficient of determination (R2) values of 0.992, 0.987 and 0.651 along with mean absolute percentage error (MAPE) values of 18.5, 2.3 and 3.7 were the outcome of ANN, FL and MLR models, respectively, when EF was predicted. Further, the re-evaluation of this study can be done in future, particularly when more data is available in the literature.