Welcome to aeon

aeon is a scikit-learn compatible toolkit for time series machine learning tasks such as classification, regression, clustering, anomaly detection, segmentation and similarity search.

  • We provide a broad library of time series algorithms, including the latest advances and state-of-the-art for many tasks.

  • Our algorithms are implemented as efficiently as possible by, for example, using numba.

  • aeon is built on top of scikit-learn, allowing for easy integration with other machine learning libraries and other time series packages.

  • We provide a range of tools for reproducing benchmarking results and evaluating time series algorithms implemented in aeon and other scikit-learn compatible packages.

Community Channels

GitHub: github.com/aeon-toolkit/aeon

Slack: aeon slack

Twitter: twitter/aeon-toolkit

LinkedIn: linkedin/aeon-toolkit

Email: contact@aeon-toolkit.org

Modules

Get started with time series classification.

Classification

Get started with time series extrinsic regression.

Regression

Get started with time series clustering.

Clustering

Get started with anomaly detection.

Anomaly Detection

Get started with forecasting

Forecasting

Get started with segmentation

Segmentation

Get started with time series transformations.

Transformations

Get started with time series distances.

Distances

Get started with time series similarity search

Similarity Search

Data structures and containers used in aeon.

Data

How to benchmark algorithms with aeon.

Benchmarking

aeon deep learning networks for time series.

Networks

Experimental Modules

Some modules of aeon are still experimental and may have changing interfaces. To support development on these modules, the deprecation policy is relaxed, so it is suggested that you integrate these modules with care. The current experimental modules are:

  • anomaly_detection

  • forecasting

  • segmentation

  • similarity_search

  • visualisation