TimesFM: Time Series Foundation Model for time-series forecasting
1–10 of 123 posts
Re: TimesFM: Time Series Foundation Model for time-series forecasting
#2"Time series" is such an over-subscribed term. What sorts of time series is this actually useful for?
For instance, will it be able to predict dynamics for a machine with thousands of sensors?
Re: TimesFM: Time Series Foundation Model for time-series forecasting
#3How can time series model be pre-trained ? I think I’m missing something.
Re: TimesFM: Time Series Foundation Model for time-series forecasting
#4How can time series model be pre-trained ? I think I’m missing something.
Third paragraph of the introduction of the mentioned paper[1] in the first paragraph of the repo.
Re: TimesFM: Time Series Foundation Model for time-series forecasting
#5Seems 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)
Re: TimesFM: Time Series Foundation Model for time-series forecasting
#6"Time series" is such an over-subscribed term. What sorts of time series is this actually useful for? For instance, will it be able to predict dynamics for a machine with thousands of sensors?
Specifically, its referring to univariate, contiguous point forecasts. Honestly, I'm a little puzzled by the benchmarks.
Re: TimesFM: Time Series Foundation Model for time-series forecasting
#7Seems 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)
What is the latency?
Re: TimesFM: Time Series Foundation Model for time-series forecasting
#8is it better than prophet from meta?
Re: TimesFM: Time Series Foundation Model for time-series forecasting
#9How can time series model be pre-trained ? I think I’m missing something.
Third paragraph of the introduction of the mentioned paper[1] in the first paragraph of the repo. [1] https://arxiv.org/abs/2310.10688
I guess they pre-trained the model to exploit common patterns found in any time-series (e.g., seasonalities, trends, etc.)... What would be interesting, though, is to see if it spots patterns that are domain-specific (e.g., the ventricular systole dip in an electrocardiogram), and possibly transfer those (that would be obviously useless in this specific example, but maybe there are interesting domain transfers out there)
Re: TimesFM: Time Series Foundation Model for time-series forecasting
#10Seems 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? :)