Artificial neural networks in the analysis of compressibility of marine clays of Grande Vitória, ES, Brazil

The compression index C_C is used to calculate the consolidation settlement of foundation on clayey soils. Several empirical relationships linking the compressibility parameters of clayey soils to their index properties have been published in the literature. This paper evaluates some of the empirical equations, which determine the Cc from laboratory index properties of soft marine clay deposits from Grande Vitória, ES (GV-ES), Southeast Brazil. Regression analyses are carried out to suggest simple correlations using both single and multiple soil parameters for estimating C_C. This paper also illustrates the potential of using Artificial Neural Networks (ANN) in the analysis of compressibility of marine clays from GV-ES. Actual laboratory test data are used in training the neural network. The results indicate that the ANN model is able to predict the compressibility of marine clays from GV-ES. It was concluded that ANN is a good alternative to the empirical equations.