Fig. 11 | Scientific Reports

Fig. 11

From: Convolutional neural network approach for automated well zonation in the Lower Bahariya member north Western Desert Egypt

Fig. 11

Comparison of expert-interpreted zonation and model-based predictions for the Shahd SE-26 well. (a) shows the traditional zonation derived from wireline log interpretation. Panels to the right illustrate the machine learning model predictions at three stages: (b) raw predictions for all window sizes, (c) results after applying the thickness filter, and (d) final filtered predictions after enforcing sequence consistency.

Back to article page