Features

SAS recently hosted its 6th annual data mining analysis and applications conference, M2003, held in Las Vegas. M2003 featured several prominent keynote speakers and 30 world-renowned data mining experts. The conference drew more than 610 attendees representing more than 20 countries around the world, and created a unique forum for exchanging new ideas and best practices.

Keynote addresses from well-known data mining thought leaders were featured on both days. Andreas Weigend, chief scientist at Amazon.com, delivered an in-depth, behind-the-scenes analysis of Amazon customer behavior and how they have successfully used the latest in data sourcing, personalization and predictive modeling to improve both the customer experience and the bottom line. Professor Lotfi Zadeh, the pioneer of "fuzzy logic" techniques and distinguished professor of computer science at the University of California, Berkeley; Tom Mitchell, professor of computer science at Carnegie Mellon University and founding director of CMU's Center for Automated Learning and Discovery; and Raghu Ramakrishnan, internationally recognized professor at the University of Wisconsin, all gave compelling talks on the latest data mining trends.

The conference featured eight breakout sessions with speakers and panel discussions highlighting creative and effective approaches that addressed issues in vertical sectors such as financial services, life sciences, retail and education. For example, David Banks, professor of statistics at Duke University, discussed new findings in Combinatorial Search in Fitting Complex Models. Banks concluded that serious model search remains difficult but that sometimes combinatorial structure in the problem can be exploited and that one can implement adaptive methods using important tools such as racing and random and restart. Randy Collica, a senior business analyst at Hewlett-Packard, discussed the latest in data mining techniques for business applications, including current data cleansing issues, data marts, "survival data mining," and integration of customer demographics with text mining. Collica concluded that text mining allows one to combine notes on customer conversations with demographics to better understand customer needs and issues, and that modeling ROI can provide up to 1000 times ROI depending on the business problem or program.

The conference website is www.sas.com/m2003