Webster West
Department of Statistics
University of South Carolina
The Data Image: A Tool for Exploring High Dimensional Data Sets
The color histogram was first introduced as a tool for visualizing higher
dimensional data by Wegman (1990). A new version of this concept, called a
data image, is discussed. Each variable is transformed into a greyscale or
color range so that a high-dimensional data set may be viewed as an image,
with observations on one axis and variables on the other. The rows and
columns of the image may be rearranged to highlight relationships between
variables. New ways of displaying the image based on various linear
orderings of a data set are discussed. The ability of the data image to
show relationships between variables and clusters in higher dimensions is
explored. A Jave applet which may be used to construct an interactive date
image via the World Wide Web is used as the basis for this exploration.
Back to Colloquium Series