Oskar Triebe

merged in #791

Hi @Hasan-Quraishi Might you be able to convert all three files to .Rmd files, and also provide their knitted html version? This would make it easier for others to reproduce...

Looking good! Do you think we could do with lower epochs (like 5) and lower regularization (like 10)?

> I did give it a try with `epoch=5` and `regularization=10`. The problem with so few epochs is that the holiday weights do not take reliable values. What do you...

@noxan LGTM! The only open question I have is: What prompted the regularization to be reduced from 10 to 0.02? Maybe we should look into standardizing the scale of the...

@noxan I was considering to drop this altogether, as it takes a long time, and as our package does not connect to the internet. Or do you think that v2...

Does prophet still do this?

Thank You @winnieshen96 for bringing this to our attention. I think this is a great suggestion. We could address this by adding documentation: - docstring in `plot_parameters`: explain meaning of...

@winnieshen96 This may not answer your question directly, but maybe this notebook can help: https://github.com/ourownstory/neural_prophet/blob/main/tutorials/feature-use/autoregression_yosemite_temps.ipynb some further resources: * https://en.wikipedia.org/wiki/Autoregressive_model * https://arxiv.org/abs/1911.12436 (section 3.2) * https://arxiv.org/abs/2111.15397 (section 2.1.3)

Hi @lleiou Great suggestion! Maybe we can expand this documentation page with more practical migration advice: https://neuralprophet.com/html/changes-from-prophet.html src: https://github.com/ourownstory/neural_prophet/blob/main/docs/source/changes-from-prophet.md Or even better, you could make a tutorial notebook on how...