Skip to content
/ CASA Public

Official implementation for "CASA: CNN Autoencoder-based Score Attention for Efficient Multivariate Long-term Time-series Forecasting" (IJCAI 2025)

Notifications You must be signed in to change notification settings

lmh9507/CASA

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CASA

The repo is the official implementation for the paper: "CASA: CNN Autoencoder-based Score Attention for Efficient Multivariate Long-term Time-series Forecasting" (IJCAI 2025)

Overall Architecture

CASA regards CNN Autoencoder-based Score Attenton improving channel-wise tokenization and shows Model Agnostic Feature including Computational Efficiency.

The main result of CASA is as the following:

Usage

  1. Install Pytorch and necessary dependencies.
pip install -r requirements.txt
  1. Train and evaluate the model. We provide all the above tasks under the folder ./scripts/. You can reproduce the results as the following examples:
# ECL dataset :  Multivariate forecasting with CASA 
bash ./scripts/long_term_forecast/ECL_script/CASA.sh

Citation

If you find this repo helpful, please cite our paper.

@misc{lee2025casacnnautoencoderbasedscore,
      title={CASA: CNN Autoencoder-based Score Attention for Efficient Multivariate Long-term Time-series Forecasting}, 
      author={Minhyuk Lee and HyeKyung Yoon and MyungJoo Kang},
      year={2025},
      eprint={2505.02011},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2505.02011}, 
}

About

Official implementation for "CASA: CNN Autoencoder-based Score Attention for Efficient Multivariate Long-term Time-series Forecasting" (IJCAI 2025)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •