Large-Scale Qualitative Modeling for System Prediction

This paper describes a software tool called GROMIT and associated methods being developed by the Statistical Sciences Group (D-1) at Los Alamos National Laboratory in support of determining dependency structures describing large system behavior. While focused on the problem of reliability prediction, the discussion is relevant to other analytic situations such as determining the actors and behaviors necessary to describe a system using agent based approaches. The paper looks at ways to improve the important qualitative reasoning which must occur in developing the structure of dependency connections, rather than on the subsequent quantitative inference or agent dynamic simulation. The focus here is on the question, “How do you know you’ve modeled the right system?” rather than on the following, related, question, “What results did we obtain from statistical or agent based modeling methods?” Contact: Dr. Andrew Koehler MS F600, D-1, Los Alamos National Laboratory Los Alamos, NM, 87545 Tel: 505-667-9513 Fax: 505-667-4470 Email: [email protected]