Time series forecasting with Greykite

Photo by Agnieszka Ziomek on Unsplash

See other stories in this series

I have been learning about time series and recently looked into Facebook Prophet and SARIMA. There is now a new Greykite library from LinkedIn that seeks to create an abstraction over multiple models including Prophet and introduces their own model called Silverkite.

I will not pretend to explain the model (because I don’t fully understand it) but LinkedIn provides the details and my interest is its application. Using the same example as previously for request rate on a hypothetical service and using Silverkite instead. The forecast fits surprisingly well!

Silverkite forecast vs observed

Also using previous Prophet and SARIMA and comparing with Silverkite, this new model fits much better! This is ofcourse exampleland with very limited training data so please keep that in mind and see the earlier post.

Compare Prophet, Sarima and Silverkite

All the code and sample data for above is in Github at this location.

See other stories in this series

If these topics interest you then reach out to me, and I will appreciate any feedback. If you would like to work on such problems you will generally find open roles as well! Please refer to LinkedIn.





Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Facial Detection using open-cv

Managing ML Training Models using ModelDB

The Intuition Behind Graph Convolutions and Message Passing

Machine Learning — a top-down view

Do you know about this powerful A.I. algorithm?


BART: Are all pretraining techniques created equal?

How To Build A Spam Classifier Using Decision Tree

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Rahul Agarwal

Rahul Agarwal


More from Medium

Practical Time Series: 1

A Comparative Analysis of different ML models for Sales Forecasting

CAN you tell the Future? Humm, Maybe …

Time-series analysis — Part 2