Time Series Machine Learning with Aeon: Classification and Regression

Matthew Middlehurst, Anthony J. Bagnall, Germain Forestier, Ali Ismail-Fawaz, Antoine Guillaume

Abstract: We present the classification and regression modules of aeon, a Python library for all machine learning tasks involving time series. aeon follows the scikit-learn API and is compatible with its utilities such as model selection and pipelines. The toolkit contains a wide range of algorithms, including the state-of-the-art and popular benchmarks for each time series learning task. We demonstrate how to use the aeon toolkit for these tasks and give an example of where these algorithms may be useful. More information and an introductory video of the toolkit modules are available on the demo webpage https://aeon-tutorials.github.io/ECML-Demo-2025/.

External IDs:dblp:conf/pkdd/MiddlehurstBFIG25