In this study, a gene expression programming (GEP) approach was employed to develop modified expressions for predicting the bearing capacity of shallow foundations founded on granular material. The model was validate against the results of load tests on full-scale and model footings obtained from the literature. Two models were developed employing different input variables in the GEP approach. The results achieved using the proposed formulae were compared with those obtained front the Meyerhof and Vesic theories. Statistical analysis was used to demonstrate that the GEP models yielded more accurate results than the traditional solutions. (C) 2015 The Japanese Geotechnical Society.
展开▼