Over the last 20 years, DePaul University in Chicago has developed a well-deserved reputation for both its degree-granting and continuing education business and technology programs. While Northwestern and the University of Chicago compete on the world academic stage, DePaul has built a critical niche educating Chicagoland's residents at the confluence of academics and practical applications. DePaul's Kellstadt Graduate School of Business part-time MBA program for working professionals is annually awarded a top ten ranking. The school's entrepreneurship programs are similarly acknowledged. And DePaul's College of Computing and Digital Media has a national reputation – and world class facilities. Finally for years, DePaul has trained hundreds of eager techies with its certification courses in such highly-pertinent areas as telecommunications/networking, project management, java and OO programming, Microsoft .Net development, relational databases/SQL and data warehousing/business intelligence. Indeed, I was an appreciative beneficiary of this training in my consulting management roles for quite some time.

One of DePaul's latest ventures, the Center for Predictive Analytics and Data Mining with the accompanying M.S. in Predictive Analytics degree, continues in this tradition. A response to the data deluge and the burgeoning demand for analytics to drive decision-making, the CPADM will open in September 2010 with focus on the applications of data analysis/machine learning across all industries, preparing students for careers in business, healthcare, utilities, education, transportation and public service.

Like many universities, DePaul doesn't support a Statistics department, instead offering data analysis and stats courses in a number of areas, most notably mathematics, economics, business and computer science. The M.S. in Predictive Analytics appears to be a collaboration of business and computer science. In fact the PA head is Chair of the Department of Marketing, while the three other dedicated faculty are from the School of Computing. This marketing and CS orientation assures an applied focus to the education.

With an obsession on business, computer science and data analysis, the applied Predictive Analytics curricula is quite different from an M.S. in Statistics, the education path of many predictive analysts, which is often highly mathematical. Prerequisites for the PA program include just two quarters of calculus, a quarter of linear algebra and a course in basic statistics. I was a bit surprised at the absence of a programming requirement. My guess is that, much like MBA applicants, target students will come to the program with a few years of post-undergraduate experience, in this case BI/analytics

Core courses for the M.S. degree include a foundation in database design, a two course sequence in data analysis from a statistical perspective, two courses in machine learning, and three in analytical marketing and CRM. Advanced elective sequences can be chosen from data mining/web analytics, visualization and imaging, database technologies, business intelligence and marketing analytics. Most of these courses are sourced from Marketing and Computer Science, reflecting the appointments of the core faculty and the applied business orientation.

The CPADM draws additional strength from its close relationship with IBM. “As the Center's first partner, IBM will provide industry expertise and guidance to professors and researchers as they apply predictive analytics and data mining skills to solve business challenges. IBM will also donate resources in the form of IBM predictive analytics software, curriculum and datasets, and provide guest lecturers. In support of the Center, IBM is also announcing its Analytic Certification in Education (ACE) program that validates a student's proficiency in IBM SPSS predictive analytics software and serves as a significant differentiator for those vying for prominent positions in today's job market.”

I'm pretty excited about the CPADM and the new M.S. program. My first reaction was one of pause, though, questioning whether the curricula would be rigorous enough mathematically. I then came to my senses, taking off the traditional statistics blinders and remembering the words of three accomplished statisticians/data analysts:

Patrick Burns – “Statistics is what stuffy professors do, I just look at my data and try to figure out what it means.”

Leo Breiman – “If our goal as a field is to use data to solve problems, then we need to move away from exclusive dependence on data models [formal statistical methods] and adopt a more diverse set of tools.”

Brad Efron – “The competition between machine learning and statistics can progress predictive science.”

Kudos to DePaul for again assuming a leadership role combining rigorous academics with a practical business focus. Expect to see this practical-academic compromise become ever more prominent, even among the most traditional elite schools, much to the benefit of both students and business. Now if we could only change CPADM's predictive analytics tool of choice from SPSS to R!

Steve Miller also blogs at Miller.OpenBI.com.