The Potential of Using Artificial Neural Networks for Prediction of Blue Nile Soil Profile in Khartoum State

Artificial Neural Networks (ANNs) are an Artificial Intelligence technique. In this study, ANNs are used for prediction of soil classification in specified locations within the flood plain of the Blue Nile in Khartoum at different depths. The study was based on the available site investigation data collected from specific areas in Sudan. About 38% of the total data collected has been used as input data. This data applied directly to neural network and the remaining percentage of the total data (about 62 % of total) has been used as tested data. Thirteen models of Neural Networks were constructed and developed to predict soil layers in specified locations in Khartoum Blue Nile area. The results were then compared with data brought from actual boreholes to check the ANN model's validity. The results indicated that Neural Networks is a useful technique for predicting the soil profile in the studied areas.