this post was submitted on 11 Mar 2024
778 points (98.9% liked)
Privacy
31991 readers
472 users here now
A place to discuss privacy and freedom in the digital world.
Privacy has become a very important issue in modern society, with companies and governments constantly abusing their power, more and more people are waking up to the importance of digital privacy.
In this community everyone is welcome to post links and discuss topics related to privacy.
Some Rules
- Posting a link to a website containing tracking isn't great, if contents of the website are behind a paywall maybe copy them into the post
- Don't promote proprietary software
- Try to keep things on topic
- If you have a question, please try searching for previous discussions, maybe it has already been answered
- Reposts are fine, but should have at least a couple of weeks in between so that the post can reach a new audience
- Be nice :)
Related communities
Chat rooms
-
[Matrix/Element]Dead
much thanks to @gary_host_laptop for the logo design :)
founded 5 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Age buckets are so archaic
I disagree, they're effective and a reasonably privacy-friendly way of predicting risk. Younger people are generally more aggressive drivers than older people, and older people generally have worse reactions than younger people. It's one of the strongest indicators for driving behavior before an infraction is recorded.
I don't like it either, but it's better imo than using one of those driving meters.
So I’m not against using age, but binning it coarsely is the issue when it can be handled much more granularly.
64-65 is probably a negligible amount of risk increase, but 64-69 is going to be much bigger. Looking at younger ages the effect is more extreme where they’re probably charging late 20’s drivers more because they’re pooled with low 20’s.
Anyway, on average it probably works out the same, but in practice I never bin data where I can avoid it, since you get better information looking at it as a continuous range.
Ah, makes sense. I'm guessing that their data sources bin ages as well, so there could be issues in moving to a continuous range.
I wish the whole thing was more transparent.
I think they totally have the computer power to use an hyper parametric model with each age as own variable. A problem this could had, is that they are not going to be enough older adults to accurately assess the risk of them and the model could end showing that 80yo's are better drivers than 30yo's.
You can use regression splines or lowess to locally weight the areas with low data based on what you do know, it keeps your parameter count down but still performs well even at the tails.