Development of Artificial Neural Networks for Prediction of Compression Coefficient of Soft Soil

Compression coefficient (Cc) of soft soil is an important parameter in solving many geotechnical problems. In this study, the main of objective is to develop an Artificial Neural Networks (ANN) for prediction of the Cc of soft soil. A total of 189 soft soil samples collected from the Ninh Binh – Hai Phong national highway project were used to carry out the laboratory tests for determining the parameters for modelling of which thirteen factors (depth of sample, clay content, moisture content, bulk density, dry density, specific gravity, void ratio, porosity, degree of saturation, liquid limit, plastic limit, plasticity index, liquidity index) were considered as input variables and the Cc was considered as a output variable for prediction models. This data was divided into two parts of training (70%) and testing (30%) datasets for building and validating the models, respectively. To validate the performance of the ANN, various methods named Mean absolute error (MAE), root mean square error (RMSE), squared correlation coefficient (R2) were used. The results show that the ANN are promising method for prediction of the Cc of soft soil. This study might help geotechnical engineers to reduce the cost of implement of laboratory tests and the time for construction.