This library is no longer actively being maintained but should work with PyTorch 1.0+

Automatic Time series forecasting

This work aims to provide a programmable interface for the automatic forecasting of time series data and implements the following papers:

• Time series forecasting with Gaussian Processes needs priors
• Automatic Construction and Natural-Language Description of Nonparametric Regression Models

Citation

If you use this library in your research, please consider citing:

@misc{chen2024autoforecaster,
  title={Autoforecaster: A library for automatic time series structure discovery and forecasting},
  author={Chen, Jonathan P and Pradhan, Neeraj and Cao, Zhen and Chandar, Siddarth and Liang, Ricky},
  journal={https://jonathanpchen.com/papers/autoforecaster.pdf},
  year={2024}
}