It almost seems like someone did a linear regression, when a logistic regression model would be more appropriate.
Comment on AI Can Tell Your Political Affiliation Just by Looking at Your Face, Researchers Find
eager_eagle@lemmy.world 5 months ago
sigh
The news:
So, in conclusion: If your face is large, you’re a conservative; if it’s skinny, you’re a liberal; and facial recognition is bad—we all know that. That seems to be all you need to know.
The paper:
Our results, suggesting that stable facial features convey a substantial amount of the signal, imply that individuals have less control over their privacy. The algorithm studied here, with a prediction accuracy of r = .22, does not allow conclusively determining one’s political views, in the same way as job interviews, with a predictive accuracy of r = .20, cannot conclusively determine future job performance.
ArtVandelay@lemmy.world 5 months ago
brian@lemmy.ca 5 months ago
What’s amusing to me is that they referred to the job interviewer having similar reliability, but didn’t say whether it was good or not. Purely let the bias of the article imply that they were highly reliable.
homesweethomeMrL@lemmy.world 5 months ago
So the headline is 100% wrong.
Everythingispenguins@lemmy.world 5 months ago
Only the words between AI and find. The rest of the headline is fine.
sugar_in_your_tea@sh.itjust.works 5 months ago
So, like only 95% is bad? That’s certainly not 100%!
ArtVandelay@lemmy.world 5 months ago
I can say with a confidence interval of 95% that the headline is bad