Several methodologies have been proposed in the past to predict the settlements of shallow foundations on granular soils. The quality of predictions was often found to be very poor. In this paper, an attempt is made to use neural network for predicting settlements of shallow foundations on granular soils and the results are encouraging, 79 settlement records with necessary foundation and soil data were collected from the literature. Out of the 79 records, 69 were used to train the neural network and the remaining ten were used to test the network. The predictions are compared with those from two traditional methods.
|Number of pages
|Transactions of the Institution of Engineers, Australia. Civil engineering
|Published - Jan 1998
ASJC Scopus subject areas
- Building and Construction
- General Engineering
- Computational Theory and Mathematics