---

Generation of missing well log data with deep learning: CNN-Bi-LSTM approach.

A Convolutional Bidirectional Long short-term memory (CNN-Bi-LSTM) network is proposed to predict missing well log data. The method successfully predicts (i) S-wave log and (ii) sonic log in the Krishna Godavari basin using the log data within the well (i) and the log data from surrounding wells (ii) respectively. The result from proposed method is compared with other methods like CNN, Bi-LSTM, ANN and Gardner’s estimation and the proposed methodology shows better results compared to others.

Study area map of KG-basin. The yellow circles indicate the NGHP-01sites used in the network training and green circles indicates the NGHP-01 sites used in prediction.

---