Have Geek Brain

Joone vs. Encog - production applications

Has anyone used Joone and Encog in production applications? Is there something you needed to do in a production application with one or the other where the product crashed?

+1


a source to share


4 answers


JOONE is apparently mostly dead, it has been on a "release candidate 1" (with a lot of bugs) since Januarl 2007. Encog has just released 2.0-beta2 and is an actively maintained product. I would go with Enkog. http://www.encog.org



+4


a source


This article might also be helpful.



http://www.dzone.com/links/choosing_an_open_source_neural_network_framework.html

+2


a source


Looked at both. Joone is quite inactive, although it has some nice design features in terms of input management and normalization. Encog is active and has a promise. These different levels of abstraction are well suited to overriding for your own purposes, although you can spend a lot of time learning the different layers of abstracted classes before getting started. Also, due to this overlap, I'm not sure the performance will be impressive, although this can be overcome by rewriting the appropriate classes for your specific application.

+1


a source


There is an article comparing all 3 available neural networks: http://www.codeproject.com/KB/recipes/xor-encog-neuroph-joone.aspx

+1


a source






More articles:


All Articles