Evaluating the Contribution of Uncertainty Management to a Fusion System

Evaluating the Contribution of Uncertainty Management to a Fusion System Report Title By its very nature, fusion requires managing uncertainty. While uncertainty management is built into many standard low-level fusion algorithms, and the importance of uncertainty management is widely recognized at all levels of the JDL hierarchy, there is less commonality of approaches to uncertainty management for high-level fusion. Evaluation of the contribution of uncertainty management to a fusion system is distinct from, although related to, evaluating the fusion system as a whole. The evaluation should consider the purpose for which fusion is being performed, and measure the contribution of uncertainty management to this purpose. The concept of expected value of sample information from decision theory provides a conceptual framework for thinking about the role of uncertainty management in fusion systems. TECHNICAL REPORT TR201312E Evaluating the Contribution of Uncertainty Management to a Fusion System Kathryn B. Laskey, Paulo C. G. Costa C4I Center George Mason University Fairfax, VA, 22030 {klaskey, pcosta} @c4i.gmu.edu Abstract— By its very nature, fusion requires managing uncertainty. While uncertainty management is built into many standard low-level fusion algorithms, and the importance of uncertainty management is widely recognized at all levels of the JDL hierarchy, there is less commonality of approaches to uncertainty management for high-level fusion. Evaluation of the contribution of uncertainty management to a fusion system is distinct from, although related to, evaluating the fusion system as a whole. The evaluation should consider the purpose for which fusion is being performed, and measure the contribution of uncertainty management to this purpose. The concept of expected value of sample information from decision theory provides a conceptual framework for thinking about the role of uncertainty management in fusion systems. By its very nature, fusion requires managing uncertainty. While uncertainty management is built into many standard low-level fusion algorithms, and the importance of uncertainty management is widely recognized at all levels of the JDL hierarchy, there is less commonality of approaches to uncertainty management for high-level fusion. Evaluation of the contribution of uncertainty management to a fusion system is distinct from, although related to, evaluating the fusion system as a whole. The evaluation should consider the purpose for which fusion is being performed, and measure the contribution of uncertainty management to this purpose. The concept of expected value of sample information from decision theory provides a conceptual framework for thinking about the role of uncertainty management in fusion systems.