Machine
intelligence
and Pattern
Analysis
Laboratory
Project
I:
Approximate Proximity for Applications in Data Mining and Visualization
Aims:
(1) a better theoretical understanding of algorithmic approximation
to exact proximity information,
(2) practical heuristics for the clustering problem in data mining,
based on techniques for approximation of proximity information and
parameterized complexity, for large spatial and categorical data
sets, and
(3) practical methods for real-time graph layout in 2D and 3D, based
on approximations to proximity information and parameterized complexity.
Colaborators:
Michael Fellows and Mike Houle
Funding:
ARC Discovery grant
|