Use ruptures to detect change points from non-stationary signals such as trend, seasonality, and variance.
With change points, you can detect anomalies or deviations from the expected behavior and gain insights into when these transitions occur.
Link to ruptures.
MLForecast: Automate External Feature Handling
Tempo: Simplified Time Series Analysis in PySpark
Chronos: Unleashing Pre-trained Language Models for Time Series Forecasting