Bayes Days at LANL
Tutorial and Workshop on Bayesian Analysis


Bayesian Reasoning in Physics: Principles and Applications

Giulio d'Agostini, University of Rome

9:00 AM - 12:30 PM, March 1-3, 1999
Physics Division Auditorium, TA-3, Bldg. 215
Los Alamos National Laboratory

In Bayesian inference, probability is the measure of the degree of belief in a hypothesis, which is a very intuitive notion. The Bayes' theorem becomes then the basic tool to evaluate the probability, combining (a priori) judgement and experimental information. This approach allows one to treat in a logically consistent way all kinds of uncertainty. The lessons will deal with uncertainty arising from measurements: inference on the value of a physics quantity from experimental observations (examining in depth the cases of observations following Gaussian, binomial and Poisson distributions); combinations of results; upper/lower limits and their combination; hypothesis tests versus probabilities of the hypotheses; systematic errors and the correlations they induce; simplified methods for routine applications (by-passing the explicit use of Bayes' theorem); type A and type B uncertainties (according to BIPM/ISO); recovery of many standard methods, but in the context of the Bayesian approach; multidimensional unfolding. (Syllabus)

The lecture notes for this course, taught in 1998 at CERN, have been finalized and issued as a CERN Yellow Report 99-03, July 1999 (vi + 175 pages) under the title ``Bayesian reasoning in high energy physics. Principles and applications''. They can be gotten from the CERN web site or from Guilio's web page.
Alternatively, the PDF files can be downloaded here
Table of Contents and Introduction,
Part 1 - Subjective probability in physics?,
Part 2 - Bayesian primer,
Part 3 - Other comments, examples, and applications,
Concluding matter - Conclusions, Acknowledgements, Bibliography, and Index.
Copies of Giulio's viewgraphs describing the six-boxes problem sixboxes.pdf (2862 KB). Sorry this PDF file is so large and cumbersome. Originally the PDF file from the visual arts folks was 28 MB! Color and all that. If someone can reduce the size of the file by some image magic, we would be glad to post it.

For other Giulio's papers, look at Giulio's home page
For a short bibliography on Bayesian inference, look at (ps, 126 KB), (pdf, 70 KB), (LaTeX BiB, 22 KB)

There is no need to preregister for this free tutorial.

Tutorial sponsored by the Enhanced Surveillance Program with refreshments kindly supplied by the ASCI Verification and Validation Program.

Workshop on Bayesian Data Analysis

1:30 PM - 4:30 PM, March 1-2, 1999
Physics Division Auditorium, TA-3, Bldg. 215
Los Alamos National Laboratory

We invite you to participate in a two-day Bayesian Data Analysis Workshop in the afternoons following the morning tutorial. The workshop will include invited talks, informal contributions, and discussion among workshop participants.

Workshop participants are encouraged to prepare short presentations describing their data analysis interests, their methodologies, and questions for discussion and debate. If you are interested in presenting your own problem, please contact the organizers in advance of the workshop.

Workshop Program

Monday
1:30 "Bayesian Estimation of Trends in the Scram Rate at Nuclear Power Plants," Harry Martz, TSA-1 (Abstract)
2:15 "Demonstration of the Role of Fluid Instabilities in Everyday Events," Kathy Prestridge, DX-3 (Abstract)
3:00 Break
3:30 Presentations of data analysis problems by attendees and discussion

Tuesday
1:30 "Bayesian Inference and Ill-Posed Inverse Problems," David Schmidt, P-21 (Abstract)
2:15 "Validation of a Hydrodynamic Code for Prediction of Mixing Instabilities," Bill Rider, XHM (Abstract)
2:45 Break
3:15 Presentations of data analysis problems by attendees and discussion


Organizers:
Ken Hanson, DX-3, 7-1402, kmh@lanl.gov or
Richard Silver, T-11, 5-1166, rns@loke.lanl.gov

If you are interested in uncertainty analysis, participate in the Uncertainty Quantification Working Group.

Return to Ken Hanson's home page.