time-series-forecasting

Model selection for time-series forecasting depends on your data patterns. This skill applies universal principles to guide decisions, suggesting ARIMA, ETS, or Theta baselines before pointing to advanced methods via references.

Backtesting a time-series forecast requires applying universal validation principles. This skill routes you to specific references that guide evaluation, validation strategies, and backtesting to ensure model reliability before deployment.

Choosing between ARIMA and ETS for time-series forecasting depends on data trend and seasonality. This skill provides routing to references that explain classical methods and helps you apply universal principles to pick the right baseline.

Building production pipelines for ML-based time-series forecasting is supported. The skill routes you to on-demand references covering deployment, ML-based approaches, and production pipeline architecture for your forecasting scenario.

No external dependencies are required. The time-series forecasting skill operates independently, using a routing framework to load scenario-driven reference files and apply universal forecasting principles without needing additional packages.

A routing framework for time-series forecasting ensures universal principles are enforced consistently. It links to on-demand references for model selection and deployment, helping teams write, review, and plan forecasts more effectively.