Machine learning (ML) is one of the most powerful yet complex tools in the toolboxes of data scientists. The difficulty of leveraging ML models will increase exponentially when the input data complexity increases. On the other hand, Earth data has become more available and accessible now. The Earth scientific research is at some level driven by the collected or simulated datasets. Using ML tools to understand the Earth is a big challenge ahead and appears to be the responsibility of Earth scientists in the next phase. However, many Earth scientists are still learning about ML and figuring how to integrate it into the existing numeric models. This session will invite speakers to talk about their experiences and share the learnt knowledge to help those who were failed or are still trying. This session calls for the presentations on a variety of ML-based Earth research topics including disruptive climate, hurricane, drought, earthquakes, human geography, socioeconomic study, agriculture, or ML-oriented cyberinfrastructure like catalog, tooling, cloud web services, high performance computing, etc. The session aims to help the community accelerate the engagement between AI and Earth data and improve our ability to deliver value-added information faster and more accurate.
View Recording
View Session Notes
View Presentations: See Session Notes above
TakeawaysTBD
