rSTSF 0.1.0

Fast, accurate and explainable time series classification through randomization

Randomized-Supervised Time Series Forest (r-STSF)

The r-STSF is a Python package for fast, accurate, and explainable time series classification. It implements the methodology described in:

Cabello, N., Naghizade, E., Qi, J., et al. Fast, accurate and explainable time series classification through randomization. Data Min Knowl Disc (2023). https://doi.org/10.1007/s10618-023-00978-w

An arXiv version of the paper is available here: Fast, Accurate and Interpretable Time Series Classification Through Randomization by Nestor Cabello, Elham Naghizade, Jianzhong Qi, and Lars Kulik.

Installation

To install r-STSF, use pip:

pip install rSTSF

Ensure you have Python 3.6 or newer installed.

Quick Start

To use r-STSF in your Python projects, import the rstsf classifier and initialize it as follows:

from rSTSF import rstsf

classifier = rstsf()

Example Usage

Below is a simple example demonstrating how to train and predict with the r-STSF classifier:

# Assuming X_train, y_train, X_test are prepared data arrays
classifier.fit(X_train, y_train)
predictions = classifier.predict(X_test)

For more detailed examples, including how to prepare your data, refer to the Jupyter notebooks provided in the GitHub repository:

• r-STSF Usage Example
• r-STSF Explainability Demo

Features

• Fast and Accurate: Achieves state-of-the-art classification accuracy efficiently, suitable for large datasets or long time series.
• Explainability: Offers insights into classification decisions, aiding understanding and trust in predictions.
• Flexible: Utilizes multiple time series representations and aggregation functions to capture diverse patterns.

License

r-STSF is licensed under the MIT License - see the LICENSE file for details.

Citation

If you use r-STSF in your research, please cite the following paper:

@article{rSTSF2023,
  title={Fast, accurate and explainable time series classification through randomization},
  author={Cabello, Nestor and Naghizade, Elham and Qi, Jianzhong and Kulik, Lars},
  journal={Data Mining and Knowledge Discovery},
  year={2023},
  publisher={Springer}
}