Week 4: The forecaster’s toolbox
Tutorial exercises
- Complete Exercise 2, 3, 10 from Section 3.7 of the book.
- Tutorial learning objectives.
- Week 4 Tutorial Solution.html
What you will learn this week
- Four benchmark forecasting methods that we will use for comparison
- Fitted values, residuals
- Forecasting using transformations
- Forecasting with decompositions
- Evaluating forecasts
Pre-seminar activities
Read Chapter 5 of the textbook and watch all embedded videos
Slides for seminar
Lectorial activities
Create a tsibble with total Holiday travellers for Victoria and Queensland from the
tourismdata set. Plot the series. What do you see?Generate 4 year ahead forecasts from all four benchmarks. Plot them using
autoplot(). Comment in the resulting forecasts.Plot the residuals from the most appropriate benchmark using
gg_tsresiduals(). What do you see?Test if the residuals are white noise. What do you conclude?
Plot point and interval forecasts from the most appropriate benchmark.
Now try a decomposition forecasting model.
Use
accuracy()to evaluate which benchmark fits the data best.Use a test set of last 3 years to check forecast accuracy.
Now use time series cross-validation to check forecast accuracy.
Exam 2024
- Section A: Q3, Q6
- Section B: Q3a, Q3b