202005181438
Honduras Face Project
tags: [ proj:face , networks , CV ]
Given a signed social network, various demographics, and faces, what interesting questions can be answered?
Existing Literature
- The Faces of Group Members Share Physical Resemblance (doi):
- similar-looking people are more likely to be friends
- not particularly surprising. the main question is one of causal direction?
- Multidimensional Homophily in Friendship Networks (link):
- this paper doesn’t actually seem that interesting. not sure what linear model they’re using, but ignoring that, what they show is that the coefficient for two variables (same sex, same ethnicity) are positive while the interaction of these two variables is negative.
- but (I’m pretty sure) all that tells you is that it’s not an additive relationship, so you get diminishing returns for homophily.
- this seems very plausible (they do mention this in the discussion). there’s redundancies involved.
- Attractiveness and Symmetry: much existing work showing that people rate averageness as more attractive
- i.e. deviation from the norm is penalized.1 except when dealing with high fashion models, where singularity is prized. this was mentioned in a recent episode on Tyler’s podcast
Backlinks
- [[master-paper-list]]
- [[honduras-face-project]]
- Nature
- [[honduras-face-project]]
- [[pediatric-transfer-learning]]
- I guess this isn’t particular to pediatric research; anytime you have some more common/majority group of individuals, and you want to study an under-represented group (or, the more stupid thing would be to think that the models trained on the majority group are somehow universal, but in fact fail completely when trained on a minority: a problem I came across in the [[honduras-face-project]]).