Application of artificial neural network for the prediction of groundwater level in hard rock region

Abstract

Artificial neural network has been shown to be an efficient tool for non-parametric modeling of data in a variety of different contexts where the output is the nonlinear function of inputs. Neural networks are the preferred tool for many predictive data mining applications because of their power, flexibility and ease of use. In the present study Feed-Forward Network based Artificial Neural Network (ANN) model is used as a method to predict the groundwater levels. The ANN model was trained using back propagated algorithm with two hidden layer, and with logsig activation function. The models are evaluated using three statistical performance criteria namely Mean Average Error (MAE), Root Mean Squared Error (RMSE) and Regression coefficient (R2). The results of the study clearly showed that ANN can be used to predict groundwater level in a hard rock region with reasonably good accuracy even in case of limited data situation. © 2011 Springer-Verlag Berlin Heidelberg.