The School of Earth and Atmospheric Sciences Presents Dr. Ghassan AlRegib, School of Electrical and Computer Engineering, Georgia Institute of Technology 

Seismic Interpretation Using Machine Learning Approaches: Recent Advances and Future Directions

In this talk, I will summarize the major and most recent advances in the field of machine learning and the most successful applications of such data-driven approaches in solving problems that relate to seismic interpretation. I will emphasize the recent interest in using deep learning for seismic interpretation tasks, such as facies classification and inversion for elastic impedance. For such applications to be successful, a large publicly available annotated datasets for training and testing models is needed. 

However, such data does not exist yet in our community. As a result, researchers have often resorted to some shortcuts that, in some cases, pose more serious challenges for the advancement of the field within our community. The talk will address such challenges and propose a few ways to overcome some of these obstacles. Specifically, the talk will focus on facies classification, structure labeling, and inversion. 

Finally, the talk will conclude with listing a few best practices to truly utilize deep learning to advance the field of seismic interpretation.