Week 5: Exponential smoothing
What you will learn this week
- Simple exponential smoothing and corresponding ETS models
- Methods with trend and corresponding ETS models
- ETS model notation
Pre-seminar activities
Read Sections 8.1-8.4 of the textbook and watch all embedded videos
Tutorial exercises
NoteTutorial Learning Objectives
- Implement and interpret exponential smoothing models.
- Able to estimate simple exponential smoothing (SES) model, interpret the smoothing parameters, and generate point and interval forecasts.
- Develop custom forecasting functions.
- Write and test user-defined functions to perform simple exponential smoothing, compare their outputs with built-in functions (like ETS()), and understand the computational steps involved.
- Optimise model parameters using numerical methods.
- Modify exponential smoothing functions to compute an error measure and use numerical optimisation methods to determine the optimal smoothing parameter
- Able to compare results with the automatic ETS estimation.
- Complete Exercises 1-4 from Section 8.8.