SIParCS 2021 - Omar Chaarawi

Omar Chaarawi

Omar Chaarawi (he/him), University of Colorado Boulder

Machine Learning Data Commons Web Portal

Recorded Talk

The Analytics and Machine Learning Group at NCAR has developed projects like machine learning tutorials and hackathons that utilize the GECKO-A dataset, but there is no introductory level tutorial that utilizes that data. Therefore, a series of introductory level ML tutorials are under development that will complement the GECKO project. The GECKO project is charged with using machine learning techniques to emulate a computationally expensive method of determining chemical concentrations present in the environment with respect to time. Tutorials involving this data represent a real-world problem where the solution is not yet necessarily refined. These tutorials will serve as a resource for climate scientists that want to take advantage of machine learning techniques, but may have knowledge gaps that impede their progress. Following these tutorials, serves as an educational resource and reflects the experience of being a climate scientist solving problems. There is opportunity to learn ML techniques and apply these to continue to solve unanswered questions. Additionally, in line with NCAR’s mission to be publicly available and support the scientific community, data and tutorials are being developed and published such that they are readily available and accessible to the general public.

Mentors: David John Gagne & John Schreck

Slides and poster