-
- In various fields of business and research it is important to be able
to visualize mutlivariate data in a way that goes beyond the standard 3D
projection. At the Electronic Visualization Laboratory (EVL) at the
University of Illinois at Chicago (UIC), research has been conducted
on visualizing multivariate functions with respect to the evaluation
of stock portfolios and interpretation of weather data. Implementation
of the first method uses the traditional technique of holding one or
more of the function's independent variables constant, therefore, reducing
the data plot dimension by one for each constant variable. This allows
the data to fit within a 3-D projection. The second area of research
implements a system of Parallel Coordinates, where each data vector is
assigned across a series of n parallel axes and connected with n-1 line
segments. This method, like most others, can easily suffer from over
plotting. One advantage is that correlations between parameters are
easily recognized by patterns between line segments.
NEXT PAGE
RETURN TO EVENTS BY YEAR MENU
Collaborators
