The Statistical Analytical System (SAS) Institute of North Carolina named SMU among the three finalists in the 2010 SAS Data Mining Shootout, a data analysis competition open to universities throughout the country.

The team will make the final presentation to the judges in quest for the winning slot on  Oct. 24 through 26 in Las Vegas in the 13th annual SAS Data Mining Conference.

Fifty universities participated in the competition, in which organizers give a problem for competing teams to find a solution using both BMI and (SAS) data mining software.

This year, teams had to analyze the medical, demographic and behavioral data of 50,788 individuals—some of whom had diabetes—in order to

determine the medical benefit of reducing the body mass indices (BMI) of a selected number of individuals by 10 percent.

Accurate selection of “high risk” individuals for such treatment leads to greater savings on future medical expenses relative to the cost of reducing BMIs of a limited number of individuals.

“We hope to make it to the top,” said Gregory Johnson, the team captain and a Ph.D. student in economics.  

“We all worked very hard on this project while working full time jobs.  We put in our hearts and soul, doing 12 hours every single day in summer,”  Johnson  said.

In order to ensure success, his team consulted previous SMU teams to learn from their experiences.

Tom Homby, faculty sponsor of the team, said data mining has been around since the 1980s, but has received increased prominence within the last five years when the software became available.  

Big companies have taken advantage of this business tool.

“Data mining involves gleaning interesting and useful facts from a large batch of information,” said Homby. “This helps companies come up with models that they can use to make business strategies that enable them to meet their financial goals.” 

He added that the future in data mining looks bright:  last October, IBM announced it would employ about 100 analysts in its Coppell office.

Manam Roy, a Ph.D. student, was in the team that participated in the project in 2008.   

She said she learned a lot during the project, which involved mining data to estimate how many rooms would be needed if flights were cancelled before 6 p.m. or after 6 p.m on a weekday.

“Companies use this tool to strategize how to react in unexpected situations in order to satisfy customers” Roy said.

Jing Ye, who was on last year’s team that also reached the finals, said she personally benefited from working with her team members. She says she got useful experience in teamwork and problem-solving.

“It was an honor for me to work with my teammates,” she added.

The competition was co-sponsored by SAS, Dow Chemical and the Central Michigan University Research cooperation.

The final order of finish will be announced Oct. 26 at the conference.