Compressive Information Extraction: A Dynamical Systems Approach
Abstract
The goal of this research was to develop a comprehensive, computationally tractable framework for addressing a broad class of problems that entail extracting information very sparsely encoded in high volume data streams. At its core was a unified vision, centered on the use of dynamical models as information encapsulators, and blending elements from dynamical systems theory, semi-algebraic geometry, sparse signal recovery and convex optimization. It included both theory development in an emerging new field compressive information extraction and an investigation of implementation issues.Relevance to the USAF mission: As emphasized in the Technology Horizons report, flexible, provably correct autonomy is a key enabler for maintaining the superiority and expanding the capabilities of the USAF in the next two decades. Autonomous systems endowed with analysis and decision capabilities can collect data, assess intention, and if necessary, take action, while at the same time substantially reducing the required manpower and cost, vis-a-vis existing unmanned vehicles.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 24, 2016
- Accession Number
- AD1004752
Entities
People
- Mario Sznaier
- Octavia Camps
Organizations
- Northeastern University