New User? Create an Account | Forgot Password?

Key Dates

Submit Session Proposal
By 1 Dec 2018
Submit Abstract/Paper
By 1 Mar 2019 By 15 Mar 2019
Author & Early Bird Reg
By 1 Apr 2019 By 15 Apr 2019

Invited Speaker

Mon - 3 Jun | 15:00 - 15:45
Tin Ka Ping Lecture Theatre (LT-1), 4/F

Professor and Chair, Department of Engineering Management and Systems Engineering George Washington University, USA

"Exploring the Use of Expert Judgment in Risk Analysis"

Risk analysis usually requires the assessment, codifications, and combination of expert judgement. There have been many theoretical models proposed and many issues debated in the open literature. However, few of the approaches developed have produced a significant database of actual expert judgement data to be analyzed with respect to its success and there is still much debate on what “success” means. The simple reason for this is that by the very nature of the requirement to use expert judgement, there is a limited amount of data to validate it.
The Delft database on expert judgement produced by multiple applications of Cooke’s Classical Model has provided an excellent vehicle for illustrating some of the issues debated in the open literature. In the Classical Model approach, experts provide percentile estimates for their uncertainty distribution for seed variables whose realization is known to the analyst but not the expert. Thus, the assessment of the quality of the expert judgement is built into the approach. From the elicited data we can provide estimates of the expert’s statistical accuracy (or calibration) and their information with respect to a specified background measure, usually the uniform distribution. These metrics are then used to establish the expert weights but can also be used to assess the “goodness” of the expert judgement.
This talk will overview some of the main themes in the expert judgement literature and use the results of the over 50 elicitation sessions with hundreds of seed variables in many different application areas gathered by users of the Classical Model to illustrate some main findings.

Dr. Thomas A. Mazzuchi received a B.A. (1978) in Mathematics from Gettysburg College, Gettysburg, PA, an M.S. (1979) and D.Sc. (1982), both in Operations Research from the George Washington University, Washington DC. Currently, he is a Professor and Chair of Engineering Management and Systems Engineering in the School of Engineering and Applied Science at the George Washington University, Washington, D.C. where he has also served as the Chair of the Operations Research Department and as Interim Dean of the School of Engineering and Applied Science.
During his academic career, he has held research contracts in development of testing procedures for both the U.S. Air Force and the U.S. Army, in spares provisioning modeling with the U. S. Postal Service, in mission assurance with NASA, and in maritime safety and risk assessment with the Port Authority of New Orleans, the Washington Office of Marine Safety, Washington State Department of Transportation, and the San Francisco Bay Area Transit Authority.
Dr. Mazzuchi’s current research interests include systems engineering processes, risk analysis, reliability growth assessment, software reliability modeling, design and inference in life testing, maintenance inspection policies, and incorporation of expert judgment into reliability and risk analysis and systems engineering methodologies.