engleski

Successfulness of inter well lithology prediction on Upper Miocene sediments with artificial neural networks

Several artificial neural networks were trained on well log curves of spontaneous potential, shallow and deep resistivity from one well for the purpose of lithology prediction in a second well. Data was taken from Upper Miocene intervals from two wells in Kloštar field. Two learning approaches and three methods of prediction were applied. Results show that the best approach for inter well lithology prediction is by training the neural network on a whole well interval. Oppositely, it was trained one neural network for each lithostratigraphic formation in Upper Miocene clastic sediments.

artificial neural networks; lithology prediction; Pannonian Basin; Miocene; Croatia

nije evidentirano