/SimTS_Representation_Learning

Primary LanguagePythonApache License 2.0Apache-2.0

SimTS: A PyTorch Implementation

This is a PyTorch implementation of the paper SimTS: Rethinking Contrastive Representation Learning for Time Series Forecasting:

@Article{SimTS2023,
    title={SimTS: Rethinking Contrastive Representation Learning for Time Series Forecasting},
    author={Zheng, Xiaochen and Chen, Xingyu and Sch{\"u}rch, Manuel and Mollaysa, Amina and Allam, Ahmed and Krauthammer, Michael},
    journal={arXiv preprint arXiv:2303.18205},
    year={2023}
}

News

  • Apr 19, 2023: Added hierarchical temporal loss from TS2Vec. Many thanks to Alexander März!
  • Apr 11, 2023: Released codes for SimTS.
More News

  • Jun 30, 1905: Albert Einstein published the theory of special relativity in Annalen der Physik.



Setup

Dependencies

A list of dependencies is provided in requirements.txt. After creating a virtual environment, we recommend installing dependencies via pip:

pip install -r /path/to/requirements.txt

Dataset

Benchmark datasets can be downloaded here.