Fall 2003

                      CS395T / CAM395T
                 "Large-Scale Data Mining"
                       M-W 4-5:30pm
                        GEO 2.102

                    Prof. Inderjit Dhillon


This is a project-oriented seminar course in data
mining, with emphasis on text data mining and bioinformatics.
Topics covered will include algorithms for (a) web search
using link analysis (Google, HITS), (b) clustering,
classification (d) factor analysis / low-rank approximations,
and (e) multidimensional visualization.
The technical tools used in the course will come from: linear
algebra, multivariate statistics, information theory and
optimization. The concepts used from these technical areas
will be covered in class, but undergraduate level linear algebra
is a pre-requisite.

A substantial portion of the course will concentrate on research
projects, where students will choose a well-defined problem.
Students will have freedom in choosing projects, as long they
are related to data mining. Projects can vary in their
theoretical/mathematical content, and in the amount of
implementation/programming involved.
Projects will be conducted by teams of up to 2-3 students.