Correlations and Copulas for Decision and Risk Analysis
Management Science
Assessing Dependence: Some Experimental Results
Management Science
Expert Systems with Applications: An International Journal
| Hi-index | 0.00 |
Large projects, especially those planned and managed by government agencies, often incur substantial cost overruns. The tolerance, particularly on the part of members of Congress, for these cost overruns has decreased, thus increasing the need for accurate, defensible cost estimates. Important aspects of creating responsible cost estimates are accounting for the uncertainties in these estimates, expressing the estimates clearly, and communicating them to decision makers. Our method for estimating cost uncertainties can be used at all stages of a project. It combines the principles of probabilistic risk analysis with procedures for expert elicitation to incorporate uncertainties and extraordinary events in cost estimates. The Department of Energy implemented this process to select a new tritium supply source. During this implementation, we identified four key issues in modeling cost risks: how to consider correlations among cost components, how to aggregate assessments of multiple experts, how to manage communication and information sharing among experts, and what is an appropriate discount rate for cost estimates.