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

TimesFM: Time Series Foundation Model for time-series forecasting

github.com

Post reply on HN
Print view View original

51–60 of 123 posts

Re: TimesFM: Time Series Foundation Model for time-series forecasting

#51

Post by zeroxfe »

whimsicalism wrote: post #17

I'm curious why we seem convinced that this is a task that is possible or something worthy of investigation. I've worked on language models since 2018, even then it was obvious why language was a useful and transferable task. I do not at all feel the same way about general univariate time series that could have any underlying process.

> I'm curious why we seem convinced that this is a task that is possible or something worthy of investigation.

There's a huge industry around time series forecasting used for all kinds of things like engineering, finance, climate science, etc. and many of the modern ones incorporate some kind of machine learning because they deal with very high dimensional data. Given the very surprising success of LLMs in non-language fields, it seems reasonable that people would work on this.

Re: TimesFM: Time Series Foundation Model for time-series forecasting

#52

Post by smokel »

aantix wrote: post #47

Would this be useful in predicting lat/long coordinates along a path? To mitigate issues with GPS drift. If not, what would be a useful model?

Map matching to a road network might be helpful here. For example, a Hidden Markov Model gives good results. See for instance this paper:

"Hidden Markov map matching through noise and sparseness" (2009)

https://www.microsoft.com/en-us/research/wp-content/uploads/...

Re: TimesFM: Time Series Foundation Model for time-series forecasting

#53

Post by wuj »

whimsicalism wrote: post #17

I'm curious why we seem convinced that this is a task that is possible or something worthy of investigation. I've worked on language models since 2018, even then it was obvious why language was a useful and transferable task. I do not at all feel the same way about general univariate time series that could have any underlying process.

Time series data are inherently context sensitive, unlike natural languages which follow predictable grammar patterns. The patterns in time series data vary based on context. For example, flight data often show seasonal trends, while electric signals depend on the type of sensor used. There's also data that appear random, like stock data, though firms like Rentech manage to consistently find unlerlying alphas. Training a multivariate time series data would be challenging, but I don't see why not for specific applications.

Re: TimesFM: Time Series Foundation Model for time-series forecasting

#54

Post by brigadier132 »

throwtappedmac wrote: post #10
nwoli wrote: post #5

Seems like a pretty small (low latency) model. Would be interesting to hook up to mouse input (x and y) and see how well it predicts where I’m gonna move the mouse (maybe with and without seeing the predicted path)

Curious George here: why are you trying to predict where the mouse is going? :)

Catching cheaters in games might seem like a good use.

Re: TimesFM: Time Series Foundation Model for time-series forecasting

#55

Post by speedgoose »

optimalsolver wrote: post #34

When it comes to time series forecasting, if the method actually works, it sure as hell isn't being publicly released.

Some times series are more predictable than others. Being good at predicting the predictable ones is useful.

For example you can easily predict the weather with descent accuracy. Tomorrow is going to be about the same than today. From there you can work on better models.

Or predicting a failure in a factory because a vibration pattern on an industrial machine always ended up in a massive failure after a few days.

But I agree that if a model is good at predicting the stock market, it’s not going to be released.

Re: TimesFM: Time Series Foundation Model for time-series forecasting

#56

Post by claytonjy »

wuj wrote: post #46

On a related note, Amazon also had a model for time series forecasting called Chronos. https://github.com/amazon-science/chronos-forecasting

And like all deep learning forecasting models thus far, it makes for a nice paper but is not worth anyone using for a real problem. Much slower than the classical methods it fails to beat.

Re: TimesFM: Time Series Foundation Model for time-series forecasting

#57

Post by yonixw »

smokel wrote: post #50
whimsicalism wrote: post #17

I'm curious why we seem convinced that this is a task that is possible or something worthy of investigation. I've worked on language models since 2018, even then it was obvious why language was a useful and transferable task. I do not at all feel the same way about general univariate time series that could have any underlying process.

The things that we are typically interested in have very clear patterns. In a way, if we find that there are no patterns, we don't even try to do any forecasting. "The Unreasonable Effectiveness of Mathematics in the Natural Sciences" [1] hints that there might be some value here. [1] https://en.m.wikipedia.org/wiki/The_Unreasonable_Effectivene...

Exactly, so for example, I think the use of this model is in cases where you want user count to have some pattern around timing. And be alerted if it has spike.

But you wouldn't want this model for file upload storage usage which only increases, where you would put alerts based on max values and not patterns/periodic values.

Re: TimesFM: Time Series Foundation Model for time-series forecasting

#58

Post by shaism »

whimsicalism wrote: post #17

I'm curious why we seem convinced that this is a task that is possible or something worthy of investigation. I've worked on language models since 2018, even then it was obvious why language was a useful and transferable task. I do not at all feel the same way about general univariate time series that could have any underlying process.

Fundamentally, the pre-trained model would need to learn a "world model" to predict well in distinct domains. This should be possible not regarding compute requirements and the exact architecture.

After all, the physical world (down to the subatomic level) is governed by physical laws. Ilya Sutskever from OpenAI stated that next-token prediction might be enough to learn a world model (see [1]). That would imply that a model learns a "world model" indirectly, which is even more unrealistic than learning the world model directly through pre-training on time-series data.

[1] https://www.youtube.com/watch?v=YEUclZdj_Sc

Post reply on HN