PI Name: Vishwas Rao
PI Institution: argonne-national-laboratory
Collaborating ANL Division: mathematics-and-computer-science-mcs

Description: In this project we will develop multiple dimensionality reduction methods for climate and weather data using machine learning techniques. Examples of such methods include but not restricted to Auto-Encoders, Variational Auto-Encoders, Convolutional Auto-Encoders, and Diffusion Maps. The output from these methods will then be used to build emulators which in turn will be used to enhance the forecasts of climate and weather models.

Testbed: amdgpu and rosie

© 2020 Joint Laboratory for System Evaluation