Usama Fayyad

Senior Researcher, Decision Theory & Adaptive Systems Group,
Microsoft Research

(for a nice propaganda blurb, check out the Researcher Profiles section at Microsoft Research homepage)

Microsoft Research, One Microsoft Way, Redmond, WA 98052-6399, USA
phone: 425-703-1528 fax: 425-936-7329

Research Interests

  • Data Mining and Knowledge Discovery in large databases
  • Clustering, especially with large data sets
  • Classification in all its types, especially things to do with decision trees
  • Statistical pattern recognition and non-linear regression
  • Machine learning theory and applications
  • Decision support systems, automated data analysis techniques, automated diagnosis systems.

Quick background:

My area of interest is data mining and knowledge discovery in databases. This means I am interested in whatever it would take to make data analysis from large databases a reality. There are plenty of research issues in this area, one that is primary is how to make data mining algorithms (techniques that originally derive from statistics, pattern recognition, learning, databases, etc.) scale to large data sets. A big problem with existing statistical analysis tools is that they assume data can be loaded to main memory, and hence run into trouble with large data sets. Another problem with classical statistical and libraries is that they require a lot of knowledge before they can be used. To see more details on what the field of data mining and KDD is all about, see pointers below or check out the Editorial for the first issue of the new journal on this topic.

For more details, check out bio on MSR's Researcher Profiles section. Before joining Microsoft Research in January 1996, I was at the Jet Propulsion Laboratory (JPL), California Institute of Technology. At JPL I headed the Machine Learning Systems Group. If you really want to find out what I'd been up to in my pre-Microsoft days, you can visit my home page at JPL.


New Journal:
Data Mining and Knowledge Discovery
Volume 1 contents: Issues 1 and 2 are both out. Issue 1 contents avaialble on-line free.
New Book:
Advances in Knowledge Discovery and Data Mining edited by U. M. Fayyad, G. Piatetsky-Shapiro, P. Smyth, and R. Uthurasamy.
A carefully edited collection describing the latest research and applications in the field of knowledge discovery in databases available now from The MIT Press.
The UW/Microsoft Research Summer Research Institute on Data Mining, July 6-11, 1997
A one week intensive workshop bringing together 75 researchers from around the world and from various disciplines including: Statistics, Databases, Systems, AI/Machine learning, KDD, Visualization, and others. See agenda, talk abstracts, reports, and more.
Communications of the ACM:
Special Issue on Data Mining: Communications of the ACM., November, 1996.
ACM Special Interest Group on Knowledge Discovery and Data Mining., FORMED! June, 1998.

Papers Online: (partial list, whatever I get around to posting so there are many missing)

OVERVIEW articles on KDD, Data Mining, and applications:

Data Mining, Scalable Algorithms:

Massive Data Sets and Statistics (issues):

Working on putting some more publication files on-line. Just give me a chance to move in and unpack physically and electronically…

Research Links of Interest:

Last Updated: 3/9/96