Information Geometry
Abstract
This is a close out memorandum report on the Information Geometry work unit. The goal of this work unit was to combine new fundamental research, which used geometric and topological analysis to develop new techniques that describe data and image complexity to make machine learning, and associated systems, "smarter" with respect to their predictive nature, and more resilient to erroneous classification. Furthermore, these techniques were evaluated for use in machine decision making in general, concentrating on multi-agents systems/Teams.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 29, 2022
- Accession Number
- AD1181603
Entities
People
- Ira S. Moskowitz
Organizations
- United States Naval Research Laboratory