Welcome to aeon¶
aeon is a scikit-learn compatible toolkit for time series tasks such as
classification, clustering, segmentation and anomaly detection.
Provides a broad library of time series algorithms, including the latest advances.
Efficient implementation of time series algorithms using
numba.Interfaces with other time series packages to provide a single framework for algorithm comparison.
Built on top of
scikit-learn, allowing for easy integration with other machine learning libraries.
Large scale changes for aeon v1.0.0
We are currently working on v1.0.0 of aeon, which includes a number of large changes to better support the maintainability of the library and the direction the developer community wish to take the package.
This includes the removal of current modules such as forecasting (to be reintroduced
under a different interface), datatypes and the legacy BaseTransformer interface. We
will also be making large changes to the base class API, removing BaseObject various
functions associated with the base module currently.
We hope these changes and the removal of long-untouched legacy code will allow aeon to grow and develop in a more sustainable way without the need for such large breaking changes in the future. Some of these changes will not come with a deprecation warning, so be wary of this when updating to v1.0.0 when it is released.
Community Channels¶
GitHub: github.com/aeon-toolkit/aeon
Slack: aeon slack
Twitter: twitter/aeon-toolkit
LinkedIn: linkedin/aeon-toolkit
Modules¶
Get started with time series classification.
Get started with time series extrinsic regression.
Get started with time series clustering.
Get started with anomaly detection.
Segmentation
Get started with time series transformations.
Get started with time series distances.
Similarity Search
Data structures and containers used in aeon.
How to benchmark algorithms with aeon.
aeon deep learning networks for time series.