- the isodata clustering algorithm as defined by Duda and
Hart (k-means clustering)
SYNOPSIS
DATASET *isodata_basic
(dset, DesiredMse)
ARGUMENTS
DATASET
*dset
The dataset which has to be clustered.
double
DesiredMse
The desired error for stopping.
RETURNS
The means of the clusters found, in DATASET format. It returns a NULL
if an error was detected.
FUNCTION
This function perfoms the isodata clustering operation as defined by Duda
and Hart, also known as k-means clustering. The clustering is perfomed on a
dataset until a desired error is reached. It returns the means of the clusters
found if the process has reached the desired error criterion. The number of
desired clusters is stored in dset->NumOutputs, i.e. dset->NumOutputs =
k.
This document was generated using api2html on Thu Mar 5 09:00:00 MET DST 1998