Customers

University of Louisville teaches data mining with SAS^®

New data mining certificate program prepares leaders

At the University of Louisville, graduate students in mathematics and computer engineering are applying SAS software to solve business problems in a wide range of career fields.

Using advanced data mining tools and concepts, the students' work includes projects that:

Study the accuracy of hospital reimbursements by diagnosis code.
Analyze pathways to success in institutional research.
Predict mobility requirements for hands-free cell phone use.
Quantitatively compare the quality of healthcare information found through various Internet sources.

"Data mining is becoming more and more important in a growing number of fields," says Patricia Cerrito, professor of mathematics at the University of Louisville. "What started as a complex process for analyzing customer data in the retail industry is now an emerging area of importance in many industry and government disciplines."

The studies mentioned above are just a smattering of the data mining applications that Cerrito's graduate students have completed in fields that range from healthcare and advertising to human resources and finance.

Cerrito, who holds a doctorate in mathematics, teaches statistics and applied mathematics at the university. Her colleague, Adel Elmaghraby, chairs the computer engineering and computer science department. Together, the mathematics and computer engineering departments have developed the university's new data mining certificate program.

According to Elmaghraby, the cross-departmental collaboration brings a depth to the program that is unique to the University of Louisville. "The core of our certificate is a combination of database and data warehousing, data handling, algorithms, and statistical methods, plus additional options in performance-related aspects and intelligent systems aspects from the computer perspective," explains Elmaghraby.

So, in addition to learning data mining techniques and statistical methods, students also learn how to structure, design and query databases for optimized data processing and analysis. "Our students will be more well-rounded since our focus is on data handling and algorithmic studies as well as statistics," says Elmaghraby. "Offering the statistical courses creates a more balanced program that is grounded in mathematics and provides a mathematical focus along with the general engineering knowledge."

Data mining skills in demand
University of Louisville's data mining certificate program satisfies a growing interest in data mining among graduate students and professionals. In fact, professionals who want to learn the science of data mining without working toward a degree can enroll in the certificate program alongside degree-seeking students who are completing the program as part of their master's degree study plans.

According to Cerrito, the program offers graduate students and working professionals important skills that provide a true return on investment for both the students and their employers. "So many businesses are overwhelmed with data, and they need employees who can extract that data and turn it into knowledge," explains Cerrito. "Today, the No. 1 job skill that employers look for is the ability to communicate technical information. And a growing need on that list is the ability to analyze and present data for business use."

Students in the new certificate program acquire both of those skills – as a result, full-time students secure good jobs upon graduation and those already in the work force advance on their chosen career paths.

To expand the program's reach even further, the university will be offering the certificate courses to online learners as well. Through a partnership with the Kentucky Virtual University, the certificate will be available through distance learning beginning in the fall of 2005. According to Cerrito, the option to take courses online is already attracting international and local students alike. "It's great to give students options in learning environments, and this distance program will provide the maximum flexibility for people in the business community who want to enhance their skills."

SAS leads data mining field
Cerrito is a strong proponent for distance learning, writing in the curriculum, and SAS software. Each year, many of her students present papers at regional SAS users group conferences and quickly comprehend why Cerrito says a strong knowledge of SAS has become a very marketable skill in the work force.

"SAS is the software that's used out in the business field, no question about it," says Cerrito. "I've looked at SAS Enterprise Miner compared to the other data mining packages, and there's ultimately no comparison. SAS Enterprise Miner offers the best integration of statistics with data mining and contains methods that other packages do not include."

In particular, Cerrito mentions link analysis, text mining and memory-based reasoning as just a few examples of where SAS excels over its competitors.

Perhaps most importantly, Cerrito and Elmaghraby say data mining teaches students to think like leaders. "The thing about data mining is that you don't approach a problem with a preset hypothesis. You start exploration, and you continue to examine the problem from different angles," says Cerrito. "In essence, that's what leadership is all about: not just answering questions, but coming up with the questions in the first place."