DL models backlog to be implemented:
| name |
TLDR |
priority |
| Autoformer |
current best $\mathcal{O}( N \log N)$ transformer. Long horizon forecasts |
|
| N-Beats |
Uncommon net architecture. Beats es-rnn(M4 winner) on M4 |
🔥 |
| N-Hits |
N-Beats predecessor. Beats Autoformer. Long horizon forecasts |
🔥 🔥 |
| TFT |
Interpretable time series forecasting. Transfromer bases. … |
DL models backlog to be implemented:
| name |
TLDR |
priority |
| Autoformer |
current best $\mathcal{O}( N \log N)$ transformer. Long horizon forecasts |
|
| N-Beats |
Uncommon net architecture. Beats es-rnn(M4 winner) on M4 |
🔥 |
| N-Hits |
N-Beats predecessor. Beats Autoformer. Long horizon forecasts |
🔥 🔥 |
| TFT |
Interpretable time series forecasting. Transfromer bases. Probabilistic forecasts |
🥶 |
| DeepAR |
RNN based. Probabilistic forecasts |
🥶 |
| NLinear |
Beats Autoformer. Simple |
🔥 |
| Vanilla Transformer |
|
🔥 |
| DeepTime |
|
🔥 |
| TIDE |
|
🔥 |
| TimesNet |
|
🔥 🔥 🔥 |
| Metrics |
|
|
| (There are few intersections or authors reused the same metrics from other papers without reproduced results) |
|
|