RI-Loss: A Learnable Residual-Informed Loss for Time Series Forecasting(RI-Loss)
December 15, 2025 ยท View on GitHub
This is the official open-source repository for our paper.
Required Packages
- pytorch==2.1.0
- tqdm==4.66.1
- python==3.8.18
- numpy==1.24.3
- matplotlib==3.7.5
Usage
1.Create ./data directory and place dataset files in ./data directory.
2.Train the model and evaluate. We provide the experiment scripts of all backbones under the folder ./scripts/. You can reproduce the results using the following commands.
cd Dlinear
# Use only MSE loss to train the model
bash ./scripts/EXP-LongForecasting/Linear/etth1.sh
# Use only RI-Loss to train the model
# In the exp_main.py file, comment out the MSE loss (lines 87 and 139) and uncomment the RI-Loss (lines 88, 89, and 140).
bash ./scripts/EXP-LongForecasting/Linear/etth1.sh
Acknowledgements
We appreciate the following GitHub repos a lot for their valuable code and efforts.
- Dlinear: https://github.com/cure-lab/LTSF-Linear
- Informer: https://github.com/zhouhaoyi/Informer2020
- Autoformer: https://github.com/thuml/Autoformer
- iTransformer: https://github.com/thuml/iTransformer
- RAFT: https://github.com/archon159/RAFT