Self-organising Map
Square Grid Problem - Demonstration Program
What does this applet do?
This simple and unsophisticated applet runs
a self-organising map tackling the square grid problem. The map has as
its input two
dimensional data (x, y), so at each node in the
map the weights vector is also two-dimensional.
As the map evolves, the weights (wxij, wyij)
at node (i, j)
are adjusted in the usual way, so that they come increasingly to resemble the
input data.
How are the data displayed?
Adjustable parameters
Two parameters can be changed. The mapsize as one
might guess, is the size of the map. The Weight adjustment
is a factor which determines the size of the adjustment to the
weights at each node every cycle. Large adjustments help stop
the map becoming trapped in local optima, but also make the
map unstable, and lead to longer convergence times.
Click here to load and run the applet.
Updated June 22, 2000.
Hugh Cartwright
(Hugh.Cartwright@chem.ox.ac.uk)
http://physchem.ox.ac.uk/~hmc/