SRA Logo

Probabilistic Risk Analysis With Hardly Any Data

A tutorial workshop
to be held in conjunction with the

Society for Risk Analysis Annual Meeting

8:00 - 5:00 pm
Sunday, 5 December 2010
(There may also be a reprise of the workshop on Thursday, 9 December 2010)

Workshop venue: Salt Palace Convention Center
Workshop hotel: Salt Lake City Marriott Downtown
Salt Lake City, Utah

This tutorial explains how you can develop a fully probabilistic risk analysis even though there may be very little empirical data available on which to base the analysis. It compares the strengths and weakness of various approaches.

Synopsis
Overview of topics
Presenters
Registration
Venue and hotel
More information
Related links

 

Synopsis

This full-day tutorial introduces and compares methods for developing a probabilistic risk analysis when little or no empirical data are available to inform the risk model. The talks are organized around the basic problems that risk analysts face: not knowing the input distributions, not knowing their correlations, not being sure about the model itself, or even which variables should be considered. Possible strategies include traditional approximative methods and recent robust and bounding methods. Numerical examples are given that illustrate the use of various methods including traditional moment propagation, PERT, maximum entropy, uniformity principle, probability bounds analysis, Bayesian model averaging and the old work horse, sensitivity analysis.

All of the approaches can be used to develop a fully probabilistic estimate useful for screening decisions and other planning. The advantages and drawbacks of the various approaches are examined. The discussion addresses how defensible decisions can be made even when little information is available, and when one should break down and collect more empirical data and, in that case, what data to collect. When properly formulated, a probabilistic risk analysis reveals what can be inferred from available information and characterizes the reliability of those inferences. In cases, where the available information is insufficient to reach dispositive conclusions, bounding probabilistic risk analysis provides a compelling argument for further empirical research and data collection.

The presentation style of the tutorial will be casual and interactive. Participants will receive a booklet of the illustrations and numerical examples used during the tutorial.




Overview of topics

Introduction

What’s wrong with qualitative risk
Fully probabilistic screening
Kinds of uncertainty
Getting something from nothing: stone soup v. blood from a stone

Don’t know the input distributions

Moments à gogo: the delta method
Moments and ranges
Put your probability in a box
Artful uniforms
Maximum entropy
PERT distributions: getting to work and gas explosions
If at first you don’t succeed
Is the quantity fixed or varying?

Interpreting outputs

Tail risks under normal theory
Chebyshev bounds on tails
Humans hate variability, but they hate uncertainty more: neuroscience of risk

Don’t know the correlations

No assumptions the Fréchet way
Dispersive Monte Carlo
Sign of the dependence
Independence is maximally entropic dependence
We don’t need no stinkin’ correlations

Not sure about the model

Three strategies for handling model uncertainty
Don’t have a risk model to start with
Don’t even know the important variables

When to break down and go get some data

Value of information and decisions under imprecise probability
Information gaps and decision thresholds
What data to get
More samples or better measurements

Conclusions

Can we agree we care about the tails?
Qualitative risk via quantitative methods
Methods in the tool chest
What should the defaults be?
Discussion



Presenters

Scott Ferson is a senior scientist at Applied Biomathematics where his research focuses on developing reliable mathematical and statistical tools for risk assessments and on methods for uncertainty analysis when empirical information is very sparse. Ferson holds a Ph.D. from the State University of New York at Stony Brook. He is author of RAMAS Risk Calc Software 4.0: Risk Assessment with Uncertain Numbers (Lewis Publishers). He has written over 100 other scholarly publications, including four other books and several software packages, in environmental risk analysis and uncertainty propagation. His research has addressed quality assurance for Monte Carlo assessments, backcalculation methods for use in remediation planning, and distribution-free methods of risk analysis appropriate for use in information-poor situations.

Jack Siegrist is a quantitative ecologist in the Ecology and Evolution Graduate Program in the Department of Ecology, Evolution and Natural Resources at Rutgers University in New Brunswick, New Jersey. He holds degrees from University of Texas at Austin and Southeastern Louisiana University. He is also a research scientist at Applied Biomathematics where his research focuses on risk perception, uncertainty comparisons, and combinatorial methods for detecting clusters in small data sets.


Registration

The registration fee is $250 by 5 November, or $300 on site. You do not need to register for the Annual Meeting to attend the workshop. Registration will be handled by

Secretariat sra@burkinc.com
Society for Risk Analysis www.sra.org
1313 Dolley Madison Boulevard, Suite 402
McLean, Virginia 22101 USA
Telephone 1-703-790-1745, Fax 1-703-790-2672



Venue

The workshop will be 8:00 - 5:00 on Sunday, 5 December 2010, at the Salt Palace Convention Center directly across the street from the Salt Lake City Marriott Downtown which is the conference hotel.

 
Hotel accommodations: Salt Lake City Marriott Downtown
Salt Lake City Marriott Downtown
75 South West Temple
Salt Lake City, Utah 84101
On-line reservations at http://cwp.marriott.com/slcut/sra/        
Phone: 1-801-531-0800 (not toll-free)
Fax: 1-801-532-4127
Toll-free: 1-800-468-3571

Workshop meeting place: Salt Palace Convention Center

The meeting room for the workshop has not yet been determined; check with the hotel concierge or convention center signage in the morning, or the night before, for directions to the meeting room.

Reserve a room at the conference hotel at the SRA rate of $119 per night (single/double occupancy) plus 6.8% tax. Be sure to mention the Society for Risk Analysis to receive the SRA group rate. This rate is also available up to three days before or after the SRA meeting, subject to availability. Reserve your room early. The SRA rate will be available until 12 November 2010, or until the SRA block of rooms is sold out. The base room rate thereafter is variable between $199 and $219, plus taxes and fees. Cancellations must be made 48 hours in advance to avoid a cancellation charge of one night's cost. You can make hotel reservations on line at http://cwp.marriott.com/slcut/sra/ or by telephoning 1-801-531-0800 or (toll-free) 1-800-468-3571. Be sure to mention the SRA when making a reservation. Reserving on-line may be considerably easier than reserving by telephone. For more information about the hotel, see the hotel fact sheet, a map of the area, or driving directions.

Weather in Salt Lake City is likely to be cold and dry in early December. The average high temperature in Salt Lake City in early December is 40 °F (5 °C) and the average low is 24 °F (-5 °C).




Thursday reprise

If there is sufficient interest, the workshop will be repeated on Thursday, 9 December 2010. To indicate your desire to attend the Thursday workshop, please email Scott Ferson at with "Thursday" in the subject line on or before 1 November 2010. We will work with interested participants to make the necessary arrangements.




More information

More information can be obtained from Scott Ferson , telephone 1-631-751-4350, fax 1-631-751–3435.




Related links

Society for Risk Analysis www.sra.org

Society for Risk Analysis Annual Meeting http://www.sra.org/events_2010_meeting.php

Hotel reservations http://cwp.marriott.com/slcut/sra/

Hotel fact sheet http://www.marriott.com/hotels/fact-sheet/travel/slcut-salt-lake-city-marriott-downtown/

Salt Palace Convention Center http://www.visitsaltlake.com/saltpalace/

Salt Lake City Wikipedia entry http://en.wikipedia.org/wiki/Salt_Lake_City

Weather in Salt Lake City http://www.wunderground.com/US/UT/Salt_Lake_City.html

The Imprecise Probabilities Project http://www.sipta.org/

Sandia National Laboratories' Epistemic Uncertainty Project http://www.sandia.gov/epistemic/

Intervals and Probability Distributions website http://ifsc.ualr.edu/jdberleant/intprob/

NSF workshop on risk perception and communication http://www.ramas.com/riskcomm.htm

 




 




 




 




 




 




 

 

More on: risk and analysis
Page transferred by Go FTP FREE Program