Data Driven Systems and Control Framework for Multiway Dynamical Systems
Abstract
The use of multiway dynamical systems is expected to providemore compact generative models than the “vectorized” counterparts, to be more robust tothe small sample size problem in data driven applications, and to provide new ways formanipulating/controlling the system by retaining and exploiting the inherent structure inthe tensorial representation. A particular emphasis ofour research will be on the data driven setting and applications where first principle modelsare not available, and on developing a computationally efficient framework associated withmultiway dynamical system identification and control.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 09, 2018
- Source ID
- FA95501810028
Entities
People
- Indika Rajapakse
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of Michigan