Geology-constrained dynamic graph convolutional networks for seismic facies classification

Ziyad Alswaidan, Motaz Alfarraj, Hamzah Luqman

Abstract: Highlights•Utilizing dynamic graph neural networks for seismic facies classification.•Added geological constraints to the network that boost the performance.•Competes with state-of-the-art architecture in facies classification.•Significantly reducing the number of parameters and the training time.