HCFA Visualization
Part of a fraud detection project conducted by the CIC-3 Data Mining
Team at Los Alamos National Laboratory. This component involves
visualization of high-dimensional clusters. 3D subsets of the
cluster dimensions can be laid out in "frames", and multiple frames
can be rendered simultaneously. The transparency of the rendered data
is inversely proportional to the Probability Density Function
estimates for the cluster. Non-linear magnification techniques are
employed to enhance visualization of the data.
Alan Keahey:
keahey@lanl.gov
