|
|
|
December 2007 Integrating Statistics in Decision Sciences
By Jamie L. Freedman There’s much more to statistics these days than crunching numbers and analyzing data. A new GW center—the Institute for Integrating Statistics in Decision Sciences—is fostering the use of modern statistical methodologies in the decision sciences. Established in January as a GW signature program with special endowment funding, the institute is directed by Refik Soyer, Ph.D. ’85, professor of decision sciences and of statistics. A cross-disciplinary enterprise, the center is housed in the Department of Decision Sciences in the School of Business, in close collaboration with the Department of Statistics in the Columbian College of Arts and Sciences and the Department of Engineering Management and Systems Engineering in the School of Engineering and Applied Science. “One of the institute’s unique characteristics is its emphasis on Bayesian thinking in decision making when dealing with uncertainty,” says Soyer, an expert in the field who has published more than 80 papers on Bayesian statistics, reliability modeling, and statistical aspects of reliability analysis. Bayesian statistics enable researchers to bring subjective input—such as expert opinion, past experience, and qualitative data—to the analysis table. “We are one of the leading institutions in the world in the application of Bayesian statistics to reliability analysis and modeling,” he emphasizes, attributing the success to the pioneering work of his mentor, Nozer Singpurwalla, professor of statistics and of decision sciences and director of GW’s Institute for Reliability and Risk Analysis. With its dual emphasis on research and education, the center is off to a flying start. “We are involved in collaborative research in a number of areas at the interface of statistics and decision sciences and are helping decision makers develop strategies and tools for dealing with uncertainty,” says Soyer. Decision sciences, he explains, focus on constructing and analyzing mathematical and statistical models of real-life problems to aid in decision making. The institute’s work has widespread applications in the real world—from helping call centers operate at peak efficiency to dealing with privacy issues in data disclosure by developing methodologies such as data camouflaging (masking data to protect people’s privacy). “Educationally, we will be developing a series of doctoral courses in decision sciences, an undergraduate course in decision modeling, and, ultimately, a new master’s degree program in business analytics,” Soyer states. The institute also plans to establish an executive education program in business analytics, as well as an MBA concentration in the field. In its first year of operation, the center launched a technical report series, spotlighting 14 publications authored or co-authored by affiliated faculty; inaugurated its seminar series with six seminars drawing colleagues from as far afield as Australia, Italy, and Turkey; and conducted a research conference on “Quality of Life” in October. In May, Soyer and his team launched a summer research grant program, supporting three recipients in projects facilitating the integration of statistics into decision sciences. “The institute is also currently providing full or partial funding to two postdoctoral fellows and four doctoral students,” says Soyer, pointing out that their research topics run the gamut from mortgage default modeling to examining computational problems in Bayesian statistics. The center’s many undertakings are overseen by the institute’s executive committee, comprising Singpurwalla; Ernest H. Forman, professor of decision sciences; Thomas A. Mazzuchi, chair and professor of engineering management and systems engineering; Srinivas Y. Prasad, chair and associate professor of decision sciences; and Murat Tarimcilar, associate dean of graduate programs at the School of Business and associate professor of decision sciences. The team also includes nine affiliated faculty members and a three-member external advisory committee, composed of prominent researchers in the field. As the institute grows, Soyer is confident that GW will play an important role in moving the field forward. “This is a great opportunity for us and for our students, who will come away from our programs very marketable,” he says. “The institute is putting GW on the map as a leading research team in Bayesian statistics and decision making.”
|
|
||
