-
**Describe the bug**
When fitting N models one for each forecasting horizon using **make_reduction** with **strategy="direct"** and **pooling="global"**, the dataset that is used to train the model…
-
- PyTorch-Forecasting version: 0.10.2
- PyTorch version1.13.0
- Python version: 3.9.15
- Operating System: Ubuntu 22.04.1 LTS
### Expected behavior
I executed TimeSeriesDataSet with a len 1 …
-
- PyTorch-Forecasting version: 0.10.3
- PyTorch version: 1.13.
- Python version: 3.9.13
- Operating System: macOS 12.6
### Expected behavior
Changing n_layers in NHiTS model to increase the n…
-
pytorch-forecasting 1.0.0
lightning 2.0.3
Python version: 3.10.9
- Operating System: iOS 13.x
I am trying to do classification with FusionTransformer. I saw in a table that it is sup…
-
-
For the API design proposal, see [this wiki entry](https://github.com/alan-turing-institute/sktime/wiki/Forecasting-API-proposal).
## Forecasters
### Atomic
- [x] NaiveForecaster (strategies={"l…
-
In sktime there are no models based on NN.
I found 3 libraries:
1) Darts
2) Python Forecasting
3) NeuralForecaster
4) Flow Forcast (2022)
5) tsai (2024)
(if there are othes, I'd like to kno…
-
Large question, and requires some discussion and then some planning. But as it keeps coming up, I think we need to discuss it:
Basic question is, should we rename the current `fmu.iteration` to `fm…
-
https://arxiv.org/abs/1910.13051
This approach feels very unique compared to the majority of forecasting models which are either using Transformers or DNNs while also being much faster.
It's for…
-
## Background
There is a rising need and opportunity for ocean management and planning that is ecologically based, including in the broadest sense of the fully integrated sociocultural and biophysic…