On the fusion of possibilistic and probabilistic information in biometric decision-making

摘要

(Svetlana Yanushkevich Special Session) Information used in decision-making in biometric systems can come from both sensors and human observers. The sensor provided information generally has a probabilistic type of uncertainty whereas human provided linguistic information typically introduces a possibilistic type of uncertainty. Here we are faced with a problem in which we must fuse information with different types of uncertainty. We provide a unified framework for the representation of these different types of information using a set measure approach for the representation of uncertain information.