Why does making larger n_forecasts change outcome for small yhats #1488
Unanswered
JCCen
asked this question in
Q&A - get help using NeuralProphet
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Hello everyone,
First off, I'd like to express my appreciation for the fantastic work on the Neural Prophet package.
I've observed an unusual behavior. When I construct a model with n_forecasts=2, it performs well. However, if I build another model with identical parameters but change n_forecasts to 150, the outcomes differ significantly.
I anticipated that the predictions from the second model at a 2 day time horizon (yhat2) would be nearly the same as the first. Yet, there's a noticeable drop in the performance of the second model for 2D horizon forecast. Could anyone shed some light on why this might be happening? Additionally, what steps could I take to mitigate this issue?
Thank you for your guidance and support.
Beta Was this translation helpful? Give feedback.
All reactions