A neural network application in integrated biostratigraphy of the Norwegian Sector, North Sea

A neural network developed at Phillips Petroleum Company helps to characterize chalk<br>reservoirs to support chalk exploration, production, and modeling. The neural network is trained<br>trom samples for which fossil counts, age, zonation, paleoenvironment, biofacies, and mode of<br>deposition have already heen determined. The neural network training is used to identify<br>inconsistencies and possible finer zonations for the training examples. The trained neural network<br>is used to make determinations for new weIl biostratigraphic data.