Reliability Engineering & System Safety · 1996 · 801 citations · 52 references
BusinessRisk ManagementRisk AnalysisRisk AssessmentUncertainty QuantificationDecision-making Under UncertaintyProbabilistic ModelingRisk EvaluationRisk ModelingUncertainty AnalysisUncertainty RepresentationEpistemic UncertaintyProbabilistic Seismic Hazard AnalysisProbabilistic Uncertainty AnalysisSeismic Hazard AnalysisUncertainty AssessmentUncertainty CharacterizationUncertainty ElicitationUncertainty In Decision Making
Epistemic uncertainty in risk analysis is difficult because encoding probability distributions from evidence and expert judgment is unavoidable. The paper examines varying levels of analytical sophistication in treating uncertainties in risk analysis and discusses when a full two‑tier uncertainty analysis is justified, highlighting potential cross‑field transfer of experience. The authors describe deterministic and probabilistic methods, outline six levels of uncertainty treatment aligned with US risk‑management evolution, and detail approaches to aggregating expert opinions for risk analysis. They illustrate these methods with a seismic‑hazard example showing how expert‑opinion aggregation can be transferred to other domains.
This paper examines different levels of analytical sophistication in the treatment of uncertainties in risk analysis, and the possibility of transfer of experience across fields of application. First, this paper describes deterministic and probabilistic methods of treatment of risk and uncertainties, and the different viewpoints that shape these analyses. Second, six different levels of treatment of uncertainty are presented and discussed in the light of the evolution of the risk management philosophy in the US. Because an in-depth treatment of uncertainties can be complex and costly, this paper then discusses when and why a full (two-tier) uncertainty analysis is justified. In the treatment of epistemic uncertainty, an unavoidable and difficult problem is the encoding of probability distributions based on scientific evidence and expert judgments. The last sections include a description of different approaches to the aggregation of expert opinions and their use in risk analysis, and a recent example of methodology and application (in seismic hazard analysis) that can be transferred to other domains.
52
Risk, Ambiguity, and the Savage Axioms
Daniel Ellsberg · The Quarterly Journal of Economics · 1961 · 7.9K citations
Business, Mathematics, Philosophy +12
Judgment Under Uncertainty: Heuristics and Biases.
Glenn Shafer, Daniel Kahnerman, Paul Slovic et al. · Journal of the American Statistical Association · 1984 · 5.7K citations
Theory of Games and Economic Behavior
E. N., John von Neumann, Oskar Morgenstern · The Journal of Philosophy · 1945 · 4.1K citations
Theory of Games and Economic Behavior
E. ROWLAND · Nature · 1946 · 3.3K citations