Content
@
0 reply
0 recast
0 reaction
Dan Romero
@dwr.eth
Welcome to @balajis.eth! He’s kindly agreed to do an AMA. Reply with your questions. :)
168 replies
54 recasts
253 reactions
horsefacts
@horsefacts.eth
I bristle a bit at the red/blue/grey model since it collapses the beautiful high dimensional complexity of a social graph into just three buckets. Simplification is useful, but encourages us to draw borders where in fact we may be interconnected. How do we find more meaningful clusters without simplifying too far?
1 reply
0 recast
4 reactions
balajis
@balajis.eth
Quantitative answer is to use spatial theory of voting to determine the actual dimensionality of the space, and how lossy the low-dimensional projection actually is. https://www.law.berkeley.edu/files/Quinn-Slides.pdf
0 reply
0 recast
9 reactions