[GRASS-user] Is "i.cluster" an implementation of the ISODATA algorithm?

[Nikos Alexandris]

Markus Neteler:

> > Since this questions comes up from time to time, it would be nice to get
> > the summary of this comparative discussion into the Wiki... volunteers
> > needed.

Nikos:
 
> Already started working it out!  If there are no objections, leave it to me.
> Please, consider assistance in reviewing it in the end, which is after all
> always useful.

I have filtered sentence by sentence the discussions, the manuals and the GRASS 
4 Image Processing Tutorial.  I am (almost) ready scratching a final version of 
a text explaining 

(not how i.cluster works -- this is already well documented, though some 
wording changes might not be bad to improve the existing manual)

the differences/similarities between the algorithm behind i.cluster and the 
ISODATA algorithm.  It might be also worthy to summarise the information in a 
table, instead of a big text -- or in addition(?).  However, I don't think 
that something like "i.cluster vs ISODATA" could be a valid title.  i.cluster 
is a module implementing some (step of an) algorithm.  ISODATA is an algorithm 
per se.

Ideas for a/the (wiki-page-)title?

- Clustering based on i.cluster
- Clustering algorithm in i.cluster
- Unsupervised clustering based on i.cluster
- Unsupervised clustering via i.cluster

Nikos