DIMACS Computational and Mathematical Epidemiology Seminar Series


Title: An Info-gap Approach to Bio- and Homeland Security

Speaker: David Fox, University of Melbourne

Date: September 7, 2006 12:00 - 1:30 pm

Location: DIMACS Center, CoRE Bldg, Room 431, Rutgers University, Busch Campus, Piscataway, NJ


Abstract:

Increasing attention is being paid to the difficult problem of evaluating risks associated with terrorist attacks. Although such attacks are increasing in frequency, they are nevertheless 'rare' events in the sense that very little data is available on which to: 1) evaluate likelihoods and 2) formulate aversion and mitigation strategies. Bayesian methods are an obvious choice for the incorporation of expert opinion although limitations in their practical application have been noted (Ouchi, 2004). An alternative framework known as Info-Gap decision theory (Ben?Haim, 2006) has been successfully applied to decision-making problems in such diverse areas as ecology (Regan et al. 2005), engineering systems (Kanno et al. 2006), and statistical inference (Fox et al, in press) Info-gap decision theory is a relatively recent development which seeks to provide immunity to extreme uncertainty in a decision-making process. This contrasts starkly with classical decision-theoretic approaches which typically aim to maximize expected utility. In this talk I will outline the essential components of an info-gap model and illustrate how an info-gap approach might assist in the evaluation of risk and uncertainty for bio-surveillance and homeland security.

References

see: DIMACS Computational and Mathematical Epidemiology Seminar Series 2006 - 2007