PREDICTION OF IMMEDIATE BEARING INDEX AND YOUNG'S MODULUS OF MODIFIED DUNE SAND WITH ARTIFICIAL NEURAL NETWORKS

Abstract

This study describes how artificial neural networks (ANNs) were used to predict the immediate bearing index (IBI) and Young's modulus of dune sand with varying blast furnace slag waste contents. The multilayer back propagation network is the most appropriate model for implementing the complexity of the nonlinear relationship between input and output parameters. It is established by incorporating a large experimental database and selecting the appropriate architecture and learning process. The ANN models proposed in this study have high applicability and reliability in predicting the young's modulus and immediate bearing index of mixes made from dune sand and blast furnace slag waste.