tdb7893
Joined 11 years ago
Karma
4,488
Submissions
1,538
Member since
Oct 2
Recent submissions
Comment · 3 days ago
I went on Reddit recently to look something up and some of the niche subreddits were decent but I was shocked at the amount of just total garbage that's there now (including lots of weird misogyny) so it's not just the bans that are a problem. It feels like every social media platform is just miserable to use now. Except occasionally posting here my only social media is just Discord servers with my actual real life friends.
Comment · 3 days ago
I'm curious if the quality of AI answers will get worse without public sources like Stack Overflow or if LLMs have gotten to a point where that doesn't matter as much.
Comment · 12 days ago
I'm doing time series forecasting for a biological system. I haven't used these foundation models but my experience is that knowledge of how things are related is baked into what variables you put into the model (which can be more than raw data, you can make any derived product you think would be useful)
but it's up to the model to figure out how they are related (if I knew exactly how they were related I wouldn't be training a forecasting model).
I'm sure someone who actually knows this stuff has a better easy to view it but my impression was that these time series foundation models aren't trained for a specific dataset but more they are trained on a procedure for how to look at datasets and make predictions. It doesn't seem to lump things into a bucket much more than any other model type (which also are just consistent procedures to look at datasets). It's not like you have radically different model types across different fields.
Comment · 12 days ago
Is crazy that production databases at places I've worked are often harder to access than this. So much stuff they do with Flock access I would be fired for if I tried it with even vaguely sensitive customer data.
Comment · 13 days ago
Somewhat more surprising to me is there's already evidence for this in baby chickens so it appears to exist even across very different animals.
Comment · 14 days ago
I definitely would not describe implementation of solar as "fanatical" by most governments, especially 20 years ago. The growth of solar is very recent, it's beat nuclear rather quickly and it's also growing very fast. A huge portion of the world's solar infrastructure is just the last 5 years.
Comment · 1 month ago
I'm always surprised to hear online that it's so bleak. I know a very wide variety of men in real life who have eventually found someone just fine on dating apps (the people I know who haven't are mostly places with a small dating pool or there's a particular reason women don't like them when they get to know them).
Comment · 1 month ago
What you're describing is a linear model, it just has an ad hoc construction instead of something like a least squares training (https://en.wikipedia.org/wiki/Linear_model). Also linear models are about linear associations and work even with non-linear predictors. Your assumption in the model is explicitly a linear association so how the predictors vary shouldn't affect it.
I don't mean to disparage amateur statistics since often there are interesting things that experts miss that amateurs have insight into but this sort of data with few data points (the data you have doesn't sound independent) and a lot of confounding factors is not trivial to accurately model. Not that everyone needs to learn statistics but this sort of data (and also this topic) probably deserves a bit more of a rigorous approach.
As a very simple example, picking a time range for the model because the model doesn't work when you go farther back adds a lot of potential bias into the model and that along with the non-independence issue (and without looking at the data myself I don't know if there are other issues) are going to lead to overconfidence in conclusions.
Comment · 1 month ago
I was checking if there was a response and upon reading your comments again, looks like your method is less a model and more just some ad hoc procedure. Based on what your conclusion is it sounds like a linear model based on demographic data would work and linear models are easy to train. Though by your summary there are probably going to be issues with the model and you have the same issues with your procedure, you just can't tell because it hasn't been rigorously specified.
There's a channel on YouTube called Statquest that teaches statistics in a pretty accessible way if you're interested in analyzing this sort of data.
Comment · 1 month ago
If you're seeing trends in your residuals that's a violation of the independence assumption of linear models (assuming you're using about a linear model since you're talking about "linear predictions").
I'm in a different field but an arbitrary 27 years (1998-2024) of data and the autocorrelation both would get flagged in a review for me. Not to give statistics homework but you should test and correct the autocorrelation issue you're describing in the model and if the data goes farther back I would go farther back, too (with how you're describing the errors going off pattern then back on it sounds like the inference here would be at least somewhat unstable depending on year chosen).
Edit: these are the assumptions and basics of how to correct for violations, in particular you're describing a violation of assumption 2 but you should test for all of them - https://www.statology.org/linear-regression-assumptions/
Comment · 1 month ago
Does this actually track over time or is it coincidence? I'm a bit confused why 1998 was chosen out of all years
Comment · 1 month ago
I've heard people say "theft" of intellectual property a lot. Also stealing an idea is common parlance. Maybe it's regional or something but I hear "theft" or similar used all the time for things other than physical goods that you lose access to.
Comment · 1 month ago
The general scam of unrealistic promises then under-delivering is such a well worn trope that while gender surely played a part in it (as it is a big part of social perceptions in general) I hesitate to blame it here. I think for anyone who has been in tech for a while it doesn't take an imagination to picture a man doing pretty much the same thing
Comment · 1 month ago
With how low many wildlife populations are, surprisingly small areas can be ecologically important
Comment · 1 month ago
Democracy works best as a diffuse institution and is more than just voting. People are supposed to have rights and power themselves and I want them to try to make things better (whether I agree with them or not) and "corporations" are just groups of people. Needing to just go with the government, especially on something so clearly within their rights, is counter to the ideals of the system.
I can understand Google somewhat here (if I was running a business I wouldn't be happy wading into a political issue like this) but the idea that they are above criticism for just falling in line still doesn't resonate with me.