Search icon CANCEL
Arrow left icon
Explore Products
Best Sellers
New Releases
Books
Videos
Audiobooks
Learning Hub
Conferences
Free Learning
Arrow right icon
Arrow up icon
GO TO TOP
Forecasting Time Series Data with Facebook Prophet

You're reading from   Forecasting Time Series Data with Facebook Prophet Build, improve, and optimize time series forecasting models using the advanced forecasting tool

Arrow left icon
Product type Paperback
Published in Mar 2021
Publisher Packt
ISBN-13 9781800568532
Length 270 pages
Edition 1st Edition
Languages
Tools
Arrow right icon
Author (1):
Arrow left icon
Greg Rafferty Greg Rafferty
Author Profile Icon Greg Rafferty
Greg Rafferty
Arrow right icon
View More author details
Toc

Table of Contents (18) Chapters Close

Preface 1. Section 1: Getting Started
2. Chapter 1: The History and Development of Time Series Forecasting FREE CHAPTER 3. Chapter 2: Getting Started with Facebook Prophet 4. Section 2: Seasonality, Tuning, and Advanced Features
5. Chapter 3: Non-Daily Data 6. Chapter 4: Seasonality 7. Chapter 5: Holidays 8. Chapter 6: Growth Modes 9. Chapter 7: Trend Changepoints 10. Chapter 8: Additional Regressors 11. Chapter 9: Outliers and Special Events 12. Chapter 10: Uncertainty Intervals 13. Section 3: Diagnostics and Evaluation
14. Chapter 11: Cross-Validation 15. Chapter 12: Performance Metrics 16. Chapter 13: Productionalizing Prophet 17. Other Books You May Enjoy

Regularizing holidays

The process of constraining a model's flexibility to help it generalize better to new data is called regularization. Chapter 4, Seasonality, featured a lengthy discussion about regularizing the effect of seasonalities in Prophet. The mathematical procedure under Prophet's hood is the same when regularizing both holiday and seasonality effects, so we can use the same concepts from the seasonality chapter and apply them to holidays.

In general, if you as the analyst find that your holidays have more control over your model than you expected, if their absolute magnitudes are higher than you believe is accurate or necessary to model your problem, then you'll want to consider regularization. Regularization will simply compress the magnitude of your holiday effects and forbid them from having as large an effect as they would otherwise. Prophet contains a holidays_prior_scale parameter to control this.

This is the same theory behind the seasonality_prior_scale...

lock icon The rest of the chapter is locked
Register for a free Packt account to unlock a world of extra content!
A free Packt account unlocks extra newsletters, articles, discounted offers, and much more. Start advancing your knowledge today.
Unlock this book and the full library FREE for 7 days
Get unlimited access to 7000+ expert-authored eBooks and videos courses covering every tech area you can think of
Renews at $19.99/month. Cancel anytime