cant do long term forecasting on yahoo stock market data about reservoirpy

[PAUL-BERNARD]

Hello,
If the predictions of your model is not good, you can try to tune the hyper-parameters ( see https://reservoirpy.readthedocs.io/en/latest/user_guide/hyper.html ).

Otherwise, I would say that predicting stock market is a very difficult task, so I'm not surprised that you have worse results than on Mackey-Glass or Lorenz.

Is ReservoirPy designed to only predict on chaotic structured data?

Reservoir computing in general is a good framework for chaotic data, but it can be used for other tasks that are not really chaotic.

Any examples of Stock Market data forecasting?

As far as I know, not with ReservoirPy, but there are a few articles here and there about using Echo State Network for this task (but I haven't read them, so I can't point specific ones).

[wilbertmatthew]

Thank you.

…

On Wed, Apr 3, 2024 at 8:11 AM PAUL BERNARD ***@***.***> wrote: Hello, If the predictions of your model is not good, you can try to tune the hyper-parameters ( see https://reservoirpy.readthedocs.io/en/latest/user_guide/hyper.html ). Otherwise, I would say that predicting stock market is a very difficult task, so I'm not surprised that you have worse results than on Mackey-Glass or Lorenz. Is ReservoirPy designed to only predict on chaotic structured data? Reservoir computing in general is a good framework for chaotic data, but it can be used for other tasks that are not really chaotic. Any examples of Stock Market data forecasting? As far as I know, not with ReservoirPy, but there are a few articles here and there about using Echo State Network for this task (but I haven't read them, so I can't point specific ones). — Reply to this email directly, view it on GitHub <#154 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAPTOI4MKOIKOBSMS3SQTWLY3PWYDAVCNFSM6AAAAABFFW3O62VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMZUGQYTAMZWGU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[neuronalX]

There are papers doing stock market prediction. One was done by Herbert Jaeger himself with students for a competition:
Iles et al (2007) Stepping forward through echoes of the past: forecasting with echo state networks
https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=368e4c95c69d3499ff1a1e46cf80b855c3b97fb5

[wilbertmatthew]

Thanks so much.

…

On Thu, Apr 4, 2024, 10:54 AM Xavier Hinaut ***@***.***> wrote: There are papers doing stock market prediction. One was done by Herbert Jaeger himself with students for a competition: Iles et al (2007) Stepping forward through echoes of the past: forecasting with echo state networks https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=368e4c95c69d3499ff1a1e46cf80b855c3b97fb5 — Reply to this email directly, view it on GitHub <#154 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAPTOI5PDA5L654JZSTYT3LY3VST3AVCNFSM6AAAAABFFW3O62VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMZXGQZTKNJWGI> . You are receiving this because you authored the thread.Message ID: ***@***.***>