Artificial neural network based production forecasting for a hydrocarbon reservoir under water injection

NEGASH Berihun Mamo,YAW Atta Dennis

Table 8 The information related to the optimum architecture of the ANN model for water production prediction.

Item Value/content Number of inputs 13 Number of hidden neuron 10 Output Water flow rate Training algorithm Bayesian regularization R2 (Training) 0.965 R2 (Testing) 0.971 Number of epochs 589 Training time 85 s