3D clustering (Hierarchical agglomerative)
- From: mohamed.rahal@xxxxxxxxx
- Date: 23 Aug 2006 07:07:18 -0700
I am interested in implementing a new approach of ascendant
agglomerative clustering, the result will be a quadtree (each node has
at most 4 children) . the visualisation will be in 3D space where the
individuals (singletons) are dispatched on a 2D (x,y) plan or surface
and the clusters in the 3rd dimension (z)
The algorithm is the following,
a. Each element of E is considered as a class and added to P.
b. Each mutual neighbor classes which can be merged in a new cluster,
among the set of classes already obtained and which have not been
merged four times, are merged in a new class and added to P.
c. The process continues until all the elements of E have been merged.
Exatelly like the HAC but in 3 dimensions
Thanks in advance for any clarification or leads to information.
Momoch.
.
- Prev by Date: Re: How to save variable names and values in text
- Next by Date: Why derivative input 1 of one block at time 1.64 is Inf or NaN?
- Previous by thread: Am I calculating error correctly? (Monte Carlo, Marquardt NL fit, etc)
- Next by thread: Why derivative input 1 of one block at time 1.64 is Inf or NaN?
- Index(es):
Relevant Pages
|
Loading
