Prediction of the Soil Compaction Parameters Using Deep Neural Networks

Estimating the optimum moisture content (OMC) and maximum dry density (MDD), or the so called compaction parameters, through laboratory tests such as Proctor test is time-consuming. This paper uses the deep neural network technique for the prediction of the soil compaction parameters for the different soil classifications in Egypt. The grain size distribution, plastic limit, and liquid limits are used as the inputs for the development of the ANNs because these variables can be easily estimated. Multiple ANNs (240 ANN) are tested, with different architectures and activation functions, in order to choose the ANN that provides the most accurate predictions. Results show that the optimum ANN that produces the best predictions consists of three hidden layers, two neurons per hidden layer, and employs the logistic activation function. This ANN provides high-accuracy results as it predicts the MDD with an R2 value of 0.864 and predicts the OMC with an R2 value of 0.924 when used on the testing set. Finally, it is shown that the deep neural network approach represents a major innovative tool for the prediction of compaction parameters as the results of this approach outperform the results of the shallow ANN that consists of a single hidden layer