Prediction of unconfined compressive strength of geopolymer stabilized clayey soil using Artificial Neural Network

Abstract Viability of Artificial Neural Network (ANN) in predicting unconfined compressive strength (UCS) of geopolymer stabilized clayey soil has been investigated in this paper. Factors affecting UCS of geopolymer stabilized clayey soil have also been reported. Ground granulated blast furnace slag (GGBS), fly ash (FA) and blend of GGBS and FA (GGBS + FA) were chosen as source materials for geo-polymerization. 28 day UCS of 283 stabilized samples were generated with different combinations of the experimental variables. Based on experimental results ANN based UCS predictive model was devised. The prediction performance of ANN model was compared to that of multi-variable regression (MVR) analysis. Sensitivity analysis employing different methods to quantify the importance of different input parameters were discussed. Finally neural interpretation diagram (NID) to visualize the effect of input parameters on UCS is also presented.