Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Low responsiveness of ML models to critical or deteriorating health conditions

nature.com

Post reply on HN
Print view View original

11–20 of 57 posts

Re: Low responsiveness of ML models to critical or deteriorating health conditions

#11

Post by Epa095 »

AlotOfReading wrote: post #8
wswope wrote: post #3

I work in the ICU monitoring field, on the R&D team of a company with live systems at dozens of hospitals and multiple FDA approvals. We use extended Kalman filters (i.e. non-blackbox "ML") to estimate certain lab values of patients that are highly indicative of them crashing, based on live data from whatever set of monitors they're hooked up to - and it's highly robust. What the authors of this paper are doing is th…

Calling EKFs "ML" is certainly a choice.

Is it less ML than linear regression?

Re: Low responsiveness of ML models to critical or deteriorating health conditions

#13

Post by thenobsta »

jvanderbot wrote: post #7
wswope wrote: post #3

I work in the ICU monitoring field, on the R&D team of a company with live systems at dozens of hospitals and multiple FDA approvals. We use extended Kalman filters (i.e. non-blackbox "ML") to estimate certain lab values of patients that are highly indicative of them crashing, based on live data from whatever set of monitors they're hooked up to - and it's highly robust. What the authors of this paper are doing is th…

Parameter estimation is ML now?

Why not? LLMs, vision models, and kalman filters all learn parameters based on data.

Re: Low responsiveness of ML models to critical or deteriorating health conditions

#14

Post by Al-Khwarizmi »

PaulHoule wrote: post #9
magicalhippo wrote: post #5

For IHM prediction, LSTM models and transformer models were trained for 100 epochs using the MIMIC-III and eICU datasets separately. I might be blind, but I don't see any mention of loss. Did they stop at 100 because it was a nice round number or because it was a good place to stop? The LSTM model they used had 7k trainable parameters, the CW-LSTM model 153k while the transformer model had 800k parameters (300k train…

What ever happened to early stopping? I see so many papers where people train neural networks with half-baked recipes. I think I saw early stopping first around 1990 but it is so often for people to pick some arbitrary number of epochs to run. I have to admit I never liked the term "early stopping", I think people should have called it just "stopping", because it makes it seem optional. Back when I was training LSTM…

I'm also annoyed about this. I suppose the main reason is because 20 years ago, if you didn't use early stopping, typically your accuracy would plummet. In earlier, smaller neural networks, overfitting was a huge issue; and lack of dropout, batch normalization, etc. made learning much more brittle.

Now the young'uns don't bother because you can just set 100 epochs, or whatever, and the result might not be optimal but it will generally be fine. Still, it's a pity because you're often wasting computational resources that could be spent in trying alternative architectures, exploring hyperparameters or whatever.

BTW, I also think "early stopping" is a terrible name. If you don't know the term, it suggests that you're going to undertrain the network, sacrificing some accuracy for efficiency. No one wants undertrained networks. I think it's not an overstatement to say that if it were called "adaptive stopping", "validation guided stopping", or even something more catchy like "smart stopping", probably more people would use it.

Re: Low responsiveness of ML models to critical or deteriorating health conditions

#15

Post by PaulHoule »

thenobsta wrote: post #13
jvanderbot wrote: post #7

Earlier quoted context omitted.

Parameter estimation is ML now?

Why not? LLMs, vision models, and kalman filters all learn parameters based on data.

A linear regression model can be written and trained as a neural net, has a loss function, all of that. Most if not all ML problems can be formulated as modelling a probability distribution

Re: Low responsiveness of ML models to critical or deteriorating health conditions

#16

Post by klodolph »

Epa095 wrote: post #11
AlotOfReading wrote: post #8

Earlier quoted context omitted.

Calling EKFs "ML" is certainly a choice.

Is it less ML than linear regression?

If you want to draw the line between ML and not ML, I think you’ll have to put Kalman filters and linear regression on the non-ML side. You can put support vector machines and neural networks on the ML side.

In some sense the exact place you draw the distinction is arbitrary. You could try to characterize where the distinction is by saying that models with fewer parameters and lower complexity tend to be called “not ML”, and models with more parameters and higher complexity tend to be called “ML”.

Re: Low responsiveness of ML models to critical or deteriorating health conditions

#17

Post by klodolph »

jvanderbot wrote: post #7
wswope wrote: post #3

I work in the ICU monitoring field, on the R&D team of a company with live systems at dozens of hospitals and multiple FDA approvals. We use extended Kalman filters (i.e. non-blackbox "ML") to estimate certain lab values of patients that are highly indicative of them crashing, based on live data from whatever set of monitors they're hooked up to - and it's highly robust. What the authors of this paper are doing is th…

Parameter estimation is ML now?

I think ML is in quotes for a reason—the reason is because the usage is not typical.

Re: Low responsiveness of ML models to critical or deteriorating health conditions

#18

Post by klodolph »

PaulHoule wrote: post #15
thenobsta wrote: post #13

Earlier quoted context omitted.

Why not? LLMs, vision models, and kalman filters all learn parameters based on data.

A linear regression model can be written and trained as a neural net, has a loss function, all of that. Most if not all ML problems can be formulated as modelling a probability distribution

That’s too reductive—ML models are statistical models. Statistical models have parameters, and in general cases, you choose the parameters with some kind of optimization algorithm.

If you play fast and loose with your definition of “ML”, you’ll end up defining it so that any statistical model is an ML model… in which case, why even bother using two different terms?

ML models are, broadly speaking, the more complicated ones with more parameters, where the behavior of the models is not really known without training. I’m sure you could nitpick to death any definition I give, but that’s fine.

Re: Low responsiveness of ML models to critical or deteriorating health conditions

#19

Post by PaulHoule »

Al-Khwarizmi wrote: post #14
PaulHoule wrote: post #9

Earlier quoted context omitted.

What ever happened to early stopping? I see so many papers where people train neural networks with half-baked recipes. I think I saw early stopping first around 1990 but it is so often for people to pick some arbitrary number of epochs to run. I have to admit I never liked the term "early stopping", I think people should have called it just "stopping", because it makes it seem optional. Back when I was training LSTM…

I'm also annoyed about this. I suppose the main reason is because 20 years ago, if you didn't use early stopping, typically your accuracy would plummet. In earlier, smaller neural networks, overfitting was a huge issue; and lack of dropout, batch normalization, etc. made learning much more brittle. Now the young'uns don't bother because you can just set 100 epochs, or whatever, and the result might not be optimal but…

I have a smart RSS reader YOShInOn which uses BERT + a probability calibrated SVM as its main model, I want to make a general purpose model trainer for text classification that is able to do harder problems.

People who hold court on ML forums will tell you fine-tuned BERT is the way to go but BERT fine-tuning doesn't seem to be compatible with early stopping with anything like the training recipes I see in the literature. Compared to old days these networks soak up knowledge like a sponge, my hunch is that with N=10,000 samples or so you don't benefit from running more than one epoch because the the network doesn't have the capacity to learn from that many samples.

I find it depressing to find arXiv papers where people copy a training recipe from other papers for BERT and compare it 5-15 different text classification problems with maybe N=500 samples. My BERT experiments take about 30 minutes so it's no small thing to do parametric scans on them, particularly when the epoch count is one of the parameters. With "smart stopping" I'm not afraid of undertraining models so I could run trainings all night and believe I'm seeing representative performance as I vary parameters.

My plan is to couple ModernBERT to a LSTM or Bi-LSTM model as the literature seems to show that this frequently ties or beats fine-tuned BERT and my experience so far as I can build reliable trainers for LSTM whereas team fined tuned BERT is indifferent to the very idea of "reliable".

Another pet peeve is all the papers with N=500 samples where I regularly get N=10,000+ in systems that I use everyday and on a rainy weekend I can lay in bed with my iPad and switch to an Android tablet when the battery runs out and get N=5000 samples. [1] When I wrote my first text classification paper we found we needed N=10,000 to get really good models, sure the world knowledge in BERT helps models learn fast and that's great (a problem I worried about in 2005 and still worry about because I think the average person wants good results at NThen there's the effort people waste with things that can't possibly work like Word2Vec, seems like people can read a lot of papers and not see it in front of them that Word2Vec is useless! I want to write a meta-analysis but instead I'm writing a diatribe and I'm not going to be happy until I repeat the paradigm with methods that are... repeatable, not for the science but for the engineering.

[1] with hallucinations as a side effect if it is a visual task but so what

Re: Low responsiveness of ML models to critical or deteriorating health conditions

#20

Post by PaulHoule »

klodolph wrote: post #18
PaulHoule wrote: post #15

Earlier quoted context omitted.

A linear regression model can be written and trained as a neural net, has a loss function, all of that. Most if not all ML problems can be formulated as modelling a probability distribution

That’s too reductive—ML models are statistical models. Statistical models have parameters, and in general cases, you choose the parameters with some kind of optimization algorithm. If you play fast and loose with your definition of “ML”, you’ll end up defining it so that any statistical model is an ML model… in which case, why even bother using two different terms? ML models are, broadly speaking, the more complicate…

I am sure there are people teach data science classes who look at it in that "reductive" way.

From the viewpoint of engineering, scikit-learn provides the same interface to linear regression that it supplies to many other models. Huggingface provides an interface to models that is similar in a lot of ways but I think a 'regression' in that it doesn't provide the bare minimum of model selection facilities needed to reliably make calibrated models. There are many problems where you could use either linear regression or a much more complex model. When it comes to "not known without training" I'm not sure how much of that is the limit of what we know right now and how much is fundamental as in the problem of "we can't really know what a computer program with free loops with do" (Halting problem) or "we can't predict what side of the side Pluto is going to be on in 30 million years" (Deterministic chaos)

(The first industrial model trainer I built was a simultaneously over and under engineered mess like most things in this industry... I didn't appreciate scikit-learn's model selection facilities and even though the data sci's I worked with had a book understanding of them, they didn't really put them to work.)

Post reply on HN