[
https://issues.apache.org/jira/browse/MATH-1267?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14734063#comment-14734063
]
Gilles commented on MATH-1267:
------------------------------
That's the basic assumption for "NeuronSquareMesh2D".
But other layouts (e.g. hexagonal) are also used. I'm not sure but I don't
think that it makes much of a difference.
Also although there are more complex maps e.g. with a variable density of
neurons.
It would be interesting to implement an "adaptive training" procedure (where
the density adapts to the training dataset). It would not be trivial to keep
it thread-safe...
Why the "Neuron2D" wrapper? A neuron is a neuron; it's the layout that adds
the concept of 1D, 2D, 3D, ...
The "Network" class manages unique identifiers.
> Grid coordinate of "Neuron" that belongs to a "NeuronSquareMesh2D"
> ------------------------------------------------------------------
>
> Key: MATH-1267
> URL: https://issues.apache.org/jira/browse/MATH-1267
> Project: Commons Math
> Issue Type: Wish
> Reporter: Gilles
> Assignee: Gilles
> Priority: Minor
> Fix For: 4.0, 3.6
>
> Attachments: LocationFinder.java
>
>
> When the network layout is a 2D grid, it useful to be able to retrieve the
> grid coordinate (row, column) of a a neuron.
> I propose to define a new class "LocationFinder" (in package
> "o.a.c.m.ml.neuralnet.twod.util") that will provide the functionality through
> a "getLocation(Neuron n)" method.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
