Fellows Lead Effort to Apply Machine Learning to Climate Change

Two Department of Energy Computational Science Graduate Fellowship recipients are leading an effort to address global climate change effects with machine-learning techniques.

Priya Donti, a third-year fellow in computer science and public policy at Carnegie Mellon University, and Kelly Kochanski, a fourth-year fellow in Earth surface processes at the University of Colorado Boulder, are on the steering committee (Donti is co-chair) for Climate Change AI. The group’s website says it is a coalition of “volunteers from academia and industry who believe in using machine learning, where it is relevant, to help tackle the climate crisis.”

Machine learning algorithms identify patterns in known data and use that information to make predictions or to classify previously unseen data. Machine learning is a key component of artificial intelligence (AI).

Climate Change AI arose from a workshop at the 2019 International Conference on Machine Learning, held at Long Beach, California, in June and a paper with 22 authors, including Donti and Kochanski. The analysis identified specific areas in which machine learning could help fill gaps in knowledge needed to address climate change impacts on electric systems, transportation and more.

The group wants to aid research at the intersection of climate change and machine learning, enable collaboration across fields and promote discussions about best practices.

For more information, listen to this podcast from writer Craig S. Smith, featuring Donti and co-chair David Rolnick of the University of Pennsylvania. A video series on machine learning applications to climate change also is available.