CNN-LSTM Model Optimized by Bayesian Optimization for Predicting Single-Well Production in Water Flooding Reservoir

Figure 2

Diagram of Conv1Ds. For the inputs sized and a Conv1D layer with filters, the outputs’ size is . Using multiple Conv1Ds, the final output size is . The model illustrated in this figure wants to predict production for the next three months, that is, it wants three outputs, which we achieve using a simple dense layer containing three neurons.