Forecasting methods are utilised in a range of industries and are important tools for both Statisticians and Operational Researchers.  This module will introduce the students to time series models and associated forecasting methods.  It will demonstrate how such models and methods can be implemented to analyse time series data, and for students to appreciate the different fields of applications.  Computer workshops will allow students to build and experiment with practical forecasting tools using data from a variety of applications.

1. Fit models for data from a large variety of sources.
2. Appreciate and use modern methods of statistical inference.
3. Forecast using a range of methods, including exponential smoothing methods and ARMA and ARIMA models.

You will be guided through learning activities appropriate to your module, which may include:

• Weekly face to face classes (e.g. labs, lectures, exercise classes)
• Electronic resources to support the learning (e.g. videos, exercise sheets, lecture notes, quizzes)

Students are also expected to undertake at least 50 hours of self-guided study throughout the duration of the module, including preparation of formative assessments.

Please see Learning Outcomes.

Time series models: decomposition, analysis and removal of trends and seasonality.

Exponential smoothing methods: single exponential, Holt and Holt-Winters methods.

Autoregressive, moving average and ARMA models.

Non-stationary series - ARIMA-models.  Forecasting using ARIMA models.