The machine learning revolution has been fueled by access to large, open and shared data. As we are well aware, this isn't always the case for geoscience. Geo-data is expensive to collect, relatively sparse and often proprietary. There is publicly available data, but it's often hidden behind opaque web interfaces and not in ML-ready form.
This project will focus on liberating datasets for the geo-hacker masses. We will discuss potential standards for documenting open geoscience datasets. We will examine the rich Poseidon seismic and petrophysical dataset, but feel free to hack on your own and join the conversation. The goal is to create notebooks, blog posts and tutorials that visualize and explore the data, ideally providing a demonstration of how machine learning can be used to extract insights. We may even send some tweets! No matter your experience level, join us and help fuel the software underground revolution!