VCformer (IJCAI 2024)

​

The repo is the official implementation for the paper: VCformer: Variable Correlation Transformer with Inherent Lagged Correlation for Multivariate Time Series Forecasting

Overall Architecture

The pseudo-code of VCformer is as simple as the following:

Usage

1. Install Python 3.8. For convenience, execute the following command.

   pip install -r requirements.txt 

2. Prepare data. You can obtain the well pre-processed datasets from [Google Drive] or [Baidu Drive], Then place the downloaded data in the folder./dataset. Here is a summary of used datasets.

3. Train and evaluate model. We provide the experiment scripts for all benchmarks under the folder ./scripts/. You can reproduce the experiment results as the following examples:

   sh ./scripts/Traffic/VCformer.sh

Citation

If you want to cite our paper, use the citation below:

@misc{yang2024vcformer,
      title={VCformer: Variable Correlation Transformer with Inherent Lagged Correlation for Multivariate Time Series Forecasting}, 
      author={Yingnan Yang and Qingling Zhu and Jianyong Chen},
      year={2024},
      eprint={2405.11470},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}

Acknowledgement

We appreciate the following Github repos a lot for their valuable code and efforts.

• Time-Series-Library (https://github.com/thuml/Time-Series-Library)
• PatchTST (https://github.com/yuqinie98/PatchTST)
• Crossformer (https://github.com/Thinklab-SJTU/Crossformer)