Grants will help advance AI techniques to address data challenges

(Nanowerk News) The U.S. Department of Energy’s (DOE) Argonne National Laboratory has received nearly $3 million in funding for two interdisciplinary projects that will further develop artificial intelligence (AI) and machine learning technology. The two grants were awarded by the DOE’s Office of Advanced Scientific Computing Research (ASCR). They will help Argonne researchers and collaborators pursue AI and machine learning work in the development of methods to tackle massive data sets or create better outcomes where little data is available.
One project is a collaboration with partners from the DOE’s Los Alamos National Laboratory, the Illinois Institute of Technology in Chicago and Johns Hopkins University. For this project, Argonne researchers will develop methods and techniques to work with very large dynamical systems. Combining mathematics and scientific principles, they will build robust and accurate surrogate models. These types of models can dramatically decrease the time and expense of running complex simulations, such as those used to predict the weather or climate.
“The models we build with this award will allow us to obtain dramatic reductions in time-to-solutions and costs,” said Romit Maulik, an Argonne computational scientist and the project lead. Maulik had been working with his colleagues to pursue these novel modeling strategies on nights and weekends. “Now we can ramp up the work we have been doing and test it on scientific use-cases right here at Argonne. For example, instead of using a massive machine to simulate the climate, we could run many smaller cheaper simulations.”
A second award went to Argonne mathematician Mihai Anitescu, who will be collaborating with Rebecca Willett at the University of Chicago. Working with an interdisciplinary team of researchers from applied mathematics, statistics and computer sciences, they will use machine learning accelerated simulations to improve forecasting, data assimilation and prediction of the frequency of extreme events.
Extreme events — such as a cascading blackout on the East Coast, a cold snap in Texas or a heatwave in Portland — can have serious consequences for people, infrastructure and the power grid. But current modeling technology is not very good at accurately predicting their frequency.
“If you look at really extreme events, you have very little data from the past to help predict what may happen in the future,” said Anitescu. In these cases, a typical machine learning approach doesn’t work. “With this project, we’re figuring out how to get around this lack of data.” This could greatly improve the ability to predict the likelihood of extreme weather events and their associated impacts on the power grid.
These two projects are part of five the DOE recently awarded for interdisciplinary work using AI to advance the science conducted in the national labs. All five are focused on developing reliable and efficient AI and machine learning methods to address a broad range of science needs.
Source: Argonne National Laboratory
Share this:
 
These articles might interest you as well: