Stress-strain modeling of sands using artificial neural networks

The authors are to be commended for proposing the use of artificial neural networks to describe the stress-strain properties of sands with different grain size distribution and stress history. However, the authors' use of a constant incremental strain value during the training and testing stages appears unjustified and restrictive to the user. In addition, the performance of the neural network may be enhanced by replacing the modeling method with the three-point scheme of Ghaboussi et al.