aeon

Why we included this project

aeon is a scikit-learn compatible Python library that covers just about every task you might run into with time-ordered data, from classification, regression, and clustering to forecasting, anomaly detection, and similarity search. Many of its algorithms are written and maintained by the researchers who originally published them, so you get faithful, current implementations of state-of-the-art methods rather than rough approximations. Teams working with sensor or wearable data, industrial monitoring, or financial and operational forecasting get a consistent API for comparing classical estimators against deep learning approaches in one place. The JMLR publication and extensive documentation also help when you need to justify your modeling choices to stakeholders.