You, Queue!April 1, 2010
Talk of the Society
This column's headline might sound like a command to a pushy bank customer during the lunchtime rush, but SIAM News readers will hear its homonym UQ---uncertainty quantification---which has rapidly grown into a major research area. A little over a decade ago, SIAM held a small workshop, with funding from the National Science Foundation and the Department of Energy, on managing uncertainty. The Chronicle of Higher Education interviewed several participants, including then SIAM president John Guckenheimer, Don Lewis, and Jim Glimm on the needs for research in the then nascent field.*
As awareness of the need for UQ grew, small numbers of minisymposia on the subject began to appear in SIAM conference programs. In recent years, UQ has become increasingly prominent at SIAM meetings, mainly (but not uniquely) at conferences in CSE. In fact, more than 35 talks in the area are scheduled for the the 2010 SIAM Annual Meeting (Pittsburgh, Pennsylvania, July 12–16).
As I write this column, UQ is at the heart of two current developments: a new solicitation from the Department of Energy's Advanced Scientific Computing Research program, and an early-April workshop at SAMSI that could lead to a SIAM activity group in this area.
The DOE solicitation, issued by the applied mathematics program (part of ASCR, which is in turn part of the Office of Science), is titled "Advanced Uncertainty (UQ) in Modeling, Simulation, and Analysis of Complex Systems." According to the solicitation,
"Uncertainty quantification refers to the broad range of activities aimed at assessing and improving confidence in simulation. There are many different sources of uncertainty and error that arise in the modeling and simulation of complex systems. For increasing the confidence of simulations, it is important to accurately characterize and quantify the effects of uncertainties and errors on mathematical models and computational algorithms."
The applied mathematics research sought for the program is related to UQ in complex systems of interest to DOE, some of which were described in an earlier DOE report, Mathematical Research Challenges in Optimization of Complex Systems (http://www.courant.nyu.edu/ComplexSystems/literature/ComplexSystemsReport.pdf). The research is further described as "UQ relevant to the simulation and analysis of complex systems on high-concurrency, extreme-scale computing architectures." A note to interested readers: Full proposals are due by April 26, 2010.
An aspect of UQ that lies at the intersection of statistics and computing is the subject of the Uncertainty Quantification Electronic Town Hall Meeting, to be held from SAMSI, the Statistical and Applied Mathematical Sciences Institute in Research Triangle Park, North Carolina, on April 6. The town hall meeting was organized in part to discuss the formation of a SIAM activity group devoted to UQ; the American Statistical Association will also consider a possible interest group in the area. One of SIAM's aims with respect to UQ is increased cooperation among statisticians and computational scientists working in the area. This town hall meeting, and the technical program scheduled for the same day, will serve as a lead-in for SAMSI's year-long (2011–12) program on methodological and practical aspects of uncertainty quantification.
I conclude by mentioning another noteworthy trend affecting SIAM conferences--the increasing number held outside the U.S. One major upcoming event, in which we hope that the entire SIAM community will participate, is the Seventh International Congress on Industrial and Applied Mathematics, which will be held in Vancouver, British Columbia, Canada, July 18–22, 2011. Prospective participants can submit proposals for minisymposia and contributed talks through the ICIAM 2011 website.
Many other international conferences will be held in the interim. One that should be of interest to many readers is the North American Meeting on Industrial and Applied Mathematics, a joint effort of SMM (Mexico), CAIMS (Canada), and SIAM. Universidad del Mar, Huatulco, in Oaxaca, Mexico, will host the meeting, which will be held December 8–10, 2010. Details will posted on the SIAM conference website as they become available.
SIAM members who are interested in an activity group in UQ, or who have general comments on the proposed activity group, are encouraged to send comments to me (email@example.com).
*"To Improve Their Models, Mathematicians Seek a ‘Science of Uncertainty,' " April 16, 1999. A sample quote from John Guckenheimer: "I don't know if I believe the predictions based on a lot of computer modeling. I don't know how much uncertainty is in the predictions, and the results are not presented in a way that makes it easy to evaluate that."
An article in Inside SIAM (SIAM News, March 2010) included the confusing suggestion that journal authors make their titles "interesting but misleading," rather than the intended "interesting but not misleading." We note further that the article, written by a well-known consultant to the publishing industry, presents general advice. Some of the information (such as the reference to a Methods section) is more pertinent to experimental sciences, but we hope that the article offers some food for thought.