I found this site by searching for 3D bezier curve force-directed graphs. In this thread mbastian says he has tried 3D bezier curves in conjunction with force-directed edge bundling, so I assume that something like this is possible.
I am interested in creating complex neural network visualizations. Let me give you a simple example to start with:

This is a medium complexity network for my lab. I positioned the layers in our neural network simulator emergent and drew the arrows using photoshop.
I am interested in coming up with a way to do this automatically. But I am even more interested in being able to plot the individual connections between neurons. This would allow us to easily see overall patterns of convergence and divergence. Consider that the network above has over ten thousand neurons and ten million connections. The ability to visualize those patterns at a glance would go a long ways towards explaining how the networks work.
I can export the positions of the individual neurons and the layers they are in as x,y,z values or as VRML and I can easily export the individual connections between the neurons. The question is, would I be able to use Gephi to create publication quality visualizations of these networks in 3D that are actually understandable by humans?
Thanks for your feedback.
Brian Mingus
Graduate student
Computational Cognitive Neuroscience Lab
University of Colorado at Boulder.Statistics:Posted by cortexman — 25 May 2011 03:10
]]>