Overview

The Greykite model

Greykite is a forecast library that allows you to do exploratory data analysis (EDA), forecast pipeline, model tuning, benchmarking, etc. It includes the Silverkite model, a forecast model developed by Linkedin, which allows feature engineering, automatic changepoint detection, holiday effects, various machine learning fitting methods, statitical prediction bands, etc. The Silverkite model has many pre-tuned templates (i.e. parameter configs) to fit for different forecast frequencies, horizons, and data patterns. Besides Silverkite, it also includes an interface for the Prophet model developed by Facebook.

This document is for users of the Greykite python library. See Simple Forecast to get started. For a more complete guide, see the rest of the tutorials in the quick start section.

Details about each model template are at Greykite models and components. You can follow those guide for advanced configuration.