DARPA SBIR ProJEct

“Koopman Operator-Based Forecasting for Nonstationary Processes from Near-Term, Limited Observational Data”

"Koopman Operator-Based Forecasting for Nonstationary Processes from Near-Term, Limited Observational Data"

Name of Contractor: AIMdyn, Inc.
Principal Investigator: Dr. Maria Fonoberova
Business Address: 1919 State St., Ste. 207, Santa Barbara, CA 93101
Phone Number: 805-687-6999
E-mail: mfonoberova@aimdyn.com

Effective Date of Contract: May 5, 2021

Sponsored by
Defense Advanced Research Projects Agency (DOD)

Issued by U.S. Army Contracting Command – Redstone
Under Contract No. W31P4Q- 21-C-0007

In this project we aim to develop prediction methodology, based on Koopman operator theory, for nonstationary processes, with low amount of observational data, performed with low computational power. The use of the Koopman operator theory enables swift adaptation to the sudden changes in the character of the received data, and, at the same time, keeping the acquired knowledge on the global time scale. We will develop the underlying algorithms and test the approach on the COVID-19 pandemic dataset. We plan to couple forecasting capability with operational aspects of logistics planning and monitoring.