Abstract

This research uses gene expression programming (GEP) in the development of a model that considers the cone penetration (CPT) data and predicts the N60 value. Some 140 CPT-SPT pairs of data points for sand, sandy, silt, and silty sand soils were used in the GEP model. The results have revealed that the GEP model can predict the standard penetration (SPT) N60 values with remarkable accuracy with high R2 values and low error for the training dataset. A comparison in terms of statistical performance between artificial neural network models (ANN) developed previously and the developed GEP model has been presented. The GEP model showed better results than the ANN model in predicting the N60 value.​