sktime/sktime

Health v2 · maintenance only

Adoption not part of the score

9965 stars · 2306 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

sktime is a unified Python framework for machine learning with time series, offering 500+ models behind a single scikit-learn-compatible API. It covers forecasting, classification, regression, clustering, anomaly and changepoint detection, transformations, pipelines, and benchmarking.

Use cases

• forecast product demand per store
• predict energy load to balance the grid
• detect anomalies in sensor data for predictive maintenance
• classify time series from wearable devices
• cluster similar time series
• detect changepoints in streaming metrics
• build and tune forecasting pipelines with AutoML
• benchmark forecasting models

When to choose

• you need one consistent API across many time series tasks and models
• you want scikit-learn-style fit/predict workflows for forecasting or classification
• you need pipelines, tuning, ensembling, and reduction for time series
• you want to swap between classical statistics, ML, and deep learning models without rewriting code

When to avoid

• you need distributed or out-of-core computation on very large datasets
• you need a GUI or CLI rather than a Python library
• you work outside time series domains
• you need real-time streaming ingestion rather than in-memory pandas/NumPy data

Facets

library · maturity stable

machine-learning data-science analytics benchmarking time-series machine-learning data-science python cross-platform windows time-series forecasting scikit-learn anomaly-detection changepoint-detection time-series-classification time-series-clustering pipelines macos linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="sktime/sktime")