Week 8: ARIMA models
What you will learn this week
- AR, MA, ARMA and ARIMA models
- Selecting model orders manually and automatically
Pre-class activities
Read Sections 9.3-9.8 of the textbook and watch all embedded videos
Slides for seminar
Workshop activities
Seminar Code
Tutorial exercises
NoteTutorial Learning Objectives
- Fit and compare ARIMA forecasting models.
- Able to use ARIMA() (with transformations where needed) to select an appropriate model, generate forecasts, and compare them against alternative specified ARIMA structures.
- Evaluate model adequacy using residual diagnostics.
- Check whether ARIMA model residuals resemble white noise (including examining autocorrelation in residuals) and use this evidence to refine or justify model choice.
- Translate between model equations and ARIMA notation.
- Recognise ARIMA model equations in algebraic form and correctly express them using ARIMA (p,d,q) notation.
Complete Exercises 7, 8, 15 (a to c), 16 (a to c) from Section 9.11 of the book.