# ATFNet

**Repository Path**: liuzl12/ATFNet

## Basic Information

- **Project Name**: ATFNet
- **Description**: No description available
- **Primary Language**: Python
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-04-30
- **Last Updated**: 2024-04-30

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# ATFNet
Official implement of ATFNet: Adaptive Time-Frequency Ensembled Network for Long-term Time Series Forecasting

## Supplementary material
### 1. Visualization of cases in Figure 1 (b) and (c)
![](figs/Figure1_showcase.png)
### 2. Full results for multi-variate forecasting under TimesNet setting. 
All baseline results are directly referenced from iTransformer[1]

[1] Liu Y, Hu T, Zhang H, et al. itransformer: Inverted transformers are effective for time series forecasting[C]//The Thirteenth International Conference on Learning Representations. 2024.
![](figs/Result.png)

## Model Architecture
![ATFNet](figs/ATFNet.png)

## Main Results
- Multi-variate long-term time series forecasting results on 8 datasets. The best results are in bold and the second best results are underlined.  

![Result](figs/Tables.png)


- Forecasting Results Visualization

![Visual](figs/ETTm2_case.svg)
<center>(Visualization of the input-192-predict-192 forecasting results on ETTm2 dataset.)</center>

## Acknowledgement
We acknowledge with great appreciation the significant contributions of the following GitHub repositories, which have provided invaluable code bases and datasets for our research.

- https://github.com/thuml/Time-Series-Library
- https://github.com/cure-lab/SCINet.
- https://github.com/aikunyi/FreTS.
- https://github.com/VEWOXIC/FITS
- https://github.com/microsoft/StemGNN/
- https://github.com/thuml/Autoformer.
- https://github.com/MAZiqing/FEDformer

---
Further detailed information will be updated in the repository subsequent to the completion of the review process.