CoST: Contrastive Learning of Disentangled Seasonal-Trend Representations for Time Series Forecasting (ICLR 2022)

Figure 1. Overall CoST Architecture.

Official PyTorch code repository for the CoST paper.

• CoST is a contrastive learning method for learning disentangled seasonal-trend representations for time series forecasting.
• CoST consistently outperforms state-of-the-art methods by a considerable margin, achieveing a 21.3% improvement in MSE on multivariate benchmarks.

Requirements

1. Install Python 3.8, and the required dependencies.
2. Required dependencies can be installed by: pip install -r requirements.txt

Data

The datasets can be obtained and put into datasets/ folder in the following way:

• 3 ETT datasets should be placed at datasets/ETTh1.csv, datasets/ETTh2.csv and datasets/ETTm1.csv.
• Electricity dataset placed at datasets/LD2011_2014.txt and run electricity.py.
• Weather dataset (link from Informer repository) placed at datasets/WTH.csv
• M5 dataset place calendar.csv, sales_train_validation.csv, sales_train_evaluation.csv, sales_test_validation.csv and sales_test_evaluation.csv at datasets/ and run m5.py.

Usage

To train and evaluate CoST on a dataset, run the script from the scripts folder: ./scripts/ETT_CoST.sh (edit file permissions via chmod u+x scripts/*).

After training and evaluation, the trained encoder, output and evaluation metrics can be found in training/<DatasetName>/<RunName>_<Date>_<Time>/.

Alternatively, you can directly run the python scripts:

python train.py <dataset_name> <run_name> --archive <archive> --batch-size <batch_size> --repr-dims <repr_dims> --gpu <gpu> --eval

The detailed descriptions about the arguments are as following:

(For descriptions of more arguments, run python train.py -h.)

Main Results

We perform experiments on five real-world public benchmark datasets, comparing against both state-of-the-art representation learning and end-to-end forecasting approaches. CoST achieves state-of-the-art performance, beating the best performing end-to-end forecasting approach by 39.3% and 18.22% (MSE) in the multivariate and univariate settings respectively. CoST also beats next best performing feature-based approach by 21.3% and 4.71% (MSE) in the multivariate and univariate settings respectively (refer to main paper for full results).

FAQs

Q: ValueError: Found array with dim 4. StandardScaler expected <= 2.

A: Please install the appropriate package requirements as found in requirements.txt, in particular, scikit_learn==0.24.1.

Q: How to set the --kernels parameter?

A: It should be list of space separated integers, e.g. --kernels 1 2 4. See the scripts folder for further examples.

Acknowledgements

The implementation of CoST relies on resources from the following codebases and repositories, we thank the original authors for open-sourcing their work.

• https://github.com/yuezhihan/ts2vec
• https://github.com/zhouhaoyi/Informer2020

Citation

Please consider citing if you find this code useful to your research.

@inproceedings{
    woo2022cost,
    title={Co{ST}: Contrastive Learning of Disentangled Seasonal-Trend Representations for Time Series Forecasting},
    author={Gerald Woo and Chenghao Liu and Doyen Sahoo and Akshat Kumar and Steven Hoi},
    booktitle={International Conference on Learning Representations},
    year={2022},
    url={https://openreview.net/forum?id=PilZY3omXV2}
}