MSc Operational Research Class

Credit Value: 3
Tutors: Lesley Walls

This class introduces students to the place of forecasting and the types of forecasting technique (short, medium and long term). It describes a variety of simple projective forecasting methods and discusses their characteristics, areas of applicability and use. These are to a certain extent ad hoc, so the section will then move onto more formal time-series decomposition and analysis using the Box-Jenkins approach; this includes some of the mathematical foundations of Box-Jenkins.

To give the class a general understanding of forecasting generally, and a toolbox of commonly-used projective forecasting techniques.

Subject specific knowledge and skills

• describe the main categories of forecasting technique, their data requirements and applicability; and describe some judgemental and associative forecasting methods
• set up short term forecasting systems, and use available data to construct and assess different models;
• explain how forecasting errors are calculated, and tracking signals calculated and used;
• decompose series using Box Jenkins analysis and draw conclusions from this.

Cognitive abilities and non-subject specific skills

• Increase ability to see patterns in numerical data.

Introduction to forecasting. Simple projective Methods (including simple adaptive smoothing). Box-Jenkins

The following may be of use but are not required:

• Forecasting: Methods and Applications (Third Edition) Makridakis, S., Wheelwright, S., and Hyndman, R.,Wiley, 1998 Book available from Amazon (UK).

One assignment on whole course.

Learning Outcomes

• main categories of forecasting technique
• different kinds of forecasting problem
• set up short term forecasting systems
• explain how forecasting errors / tracking signals calculated;
• decompose series using Box Jenkins analysis and draw conclusions
• Increase ability to see patterns in numerical data.