Session 3

summary: Simple forecasting methods presented in the last session e.g. Naive, are not able to capture trend and seasonality in the past data. In this session , you will learn how to use expoentnial smoothing to capture both trend and seasonality in time series and use it to forecast the future.

Questions

  1. Which features of time series are captured by Exponential Smoothing?
  2. What are different forecasting approaches in Exponential Smoothing?
  3. How Exponential Smoothing combines different features of time series?
  4. What is the difference between an Exponential Smoothing method and an Exponential Smoothing model?
  5. What are different Smoothing parameters in Exponential Smoothing? and How they control the rate of changes in systematic patterns over time?
  6. How to estimate parameters of Exponential Smoothing models?
  7. How to specify and estimate Exponential Smoothing models using R functions in fable?

Schedule

Session 03 slides: ETS: Exponential Smoothing models
Lab Rmarkdown: Fitting ETS models and evaluate their forecast accuracy

Readings

📖 Exponential Smoothing models: Chp 8, Exponential smoothing Required
📖 Exponential Smoothing models: smooth package optional

Demo

📖 Shiny App: Exponential Smoothing models Demo optional