Is it possible to predict multiple timesteps? #369
Replies: 5 comments 2 replies
-
|
Multiple timesteps into the future as in, given a(x,t), predict u(x,t+2)? Or as in given a(x,t) predict u(x,t) and u(x,t+1) at the same time? |
Beta Was this translation helpful? Give feedback.
-
|
Given an input T, predict T+1, then use it as the input to predict T+2 and so on. Hopefully that makes sense |
Beta Was this translation helpful? Give feedback.
-
|
The library doesn't currently have a training module written for the autoregressive use case |
Beta Was this translation helpful? Give feedback.
-
|
https://github.com/neuraloperator/neuraloperator/blob/master/fourier_2d_time.py This might be what you're looking for. |
Beta Was this translation helpful? Give feedback.
-
|
The link is now broken. Can you help me finding it? |
Beta Was this translation helpful? Give feedback.
-
|
Is it possible? |
Beta Was this translation helpful? Give feedback.
-
|
Yes. You can either do auto-regressive prediction and feed your model's prediction back to it as input, or you can use a 3D FNO to predict a spatio-temporal volume. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I'm currently using the library to make time step based predictions. Does the library have the capability to predict multiple time steps into the future? (T+2,T+3...) or am i just missing something
Beta Was this translation helpful? Give feedback.
All reactions