The application of category theory and analysis of receiver operating characteristics to information fusion.

摘要

This research provides a mathematical foundation for the science of fusing information from automatic target recognition and classification systems. It also provides a method for selecting the best classification system from a family of systems for any finite number of classes. The research will allow Air Force and DoD engineers to design better ATR systems at reduced cost as well as allowing Air Force testing and acquisition agencies to better write requirements, compete classification systems, and select the best operating parameters of the systems we use.