Week 9.5 Funonesense
Visualizing high dimensional space.
https://www.youtube.com/watch?v=wvsE8jm1GzE
I was struck when I saw this video. It reminded me of our discussion of the dimensionality of niches and/or guilds. Somehow, I'm not quite sure how, this program clusters things together in 3D space based on their nth dimensional relationship to each other. I was wondering, if we defined a bunch of species by 4-15 significant dimensions (body size, diet, motility, metabolism, body plan) as numbers and plugged them into this thing as datapoints, could this program visualize them into 'real' ecological groups? Do you guys think there's some reason why this wouldn't work?
I think neural networks are cool, but I don't understand them.
Comments
Post a Comment