This event has ended. Create your own event on Sched.
Data to Action: Increasing the Use and Value of Earth Science Data and InformationFor 20 years, ESIP meetings have brought together the most innovative thinkers and leaders around Earth observation data, thus forming a community dedicated to making Earth observations more discoverable, accessible and useful to researchers, practitioners, policymakers, and the public.

The ESIP Summer Meeting has already taken place, but check out the ESIP Summer Meeting Highlights Webinar: https://youtu.be/vbA8CuQz9Rk.
Back To Schedule
Thursday, July 18 • 1:30pm - 3:00pm
An ESIP community's working session on machine learning: introducing adoptable use cases and beyond

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
The large volume of freely available Earth and environmental data and the fast developing computational capacity have powered the rapid growth of discovery and exploration using machine learning (ML) within Earth science information community.

Based on a community survey conducted by ESIP Machine Learning Cluster in 2018, the majority of the participants expressed the desire of introductory materials for ML applications as well as the interests in curated ML datasets. With this in mind, the ML cluster has since then aligned our efforts on 1) generating Earth science specific ML application examples and 2) curating ML data repositories for ESIP community.
In this working session, we invite everyone in the ESIP community who is interested in various ML applications to contribute to a discussion on how to move these efforts forward in a most efficient and community-driven manner. The session will begin with presentations from members of the cluster on a) the development of sample use cases for learning about ML, 2) the curation of a centralized metadata repository of open source ML suitable data sets from various repositories, and 3) a demo of Data Driven Discovery of Models (D3M) for composing ML pipelines to simultaneously solve various problems. The session will conclude with an open discussion on the cluster’s priority and efforts for the following 6-month to best serve the ESIP community.

We encourage all ESIP ML enthusiastics to join our session and contribute to the cluster’s initiatives.
  • Yuhan Rao (University of Maryland, College Park): Machine learning applications using R and open source earth science datasets
    • In this presentation, I will share the status of the efforts on developing sample ML applications using open source earth science data sets (from UCI repository) and ML package in R. The use case can be adopted by beginners to start their own applications.
  • Arif Albayrak (NASA Goddard Space Flight Center): Enabling machine learning applications in earth science community through curated open source data sets
    • This presentation features the cluster’s initiative to gather available open source data sets which are suitable for ML applications in earth science. The goal of this effort is to enable future development of ML applications for users with different level of experiences.
  • Sujen Shah (NASA Jet Propulsion Laboratory): Demonstration of automated machine learning application pipeline through DARPA Data Driven Discovery of Models (D3M)

View the Recording on YouTube

avatar for Anne Wilson

Anne Wilson

Senior Software Engineer, Laboratory for Atmospheric and Space Physics

Thursday July 18, 2019 1:30pm - 3:00pm PDT
Room 316
  Room 316, Working Session