The process of forecasting is one that has been applied by many and has evolved significantly over
the past several decades. In the foodservice industry, effective and profitable operation relies
heavily on accurate and reliable sales forecasts. This project was conducted with the aim of
determining if it is possible to develop a forecasting tool consisting of a model or selection of
models capable of forecasting daily sales for different restaurants with similar operating
environments which can subsequently be developed into an Expert System. This Expert System
should replicate the expertise, comprehension, intuition, and intelligence of the forecast knowledge
expert in the process of forecasting daily sales in the foodservice industry and can then be used by
individuals who has limited or no experience/understanding of the process of forecasting.
This document includes a background on the company for which the model will be developed. It
also highlights the problem that were addressed by this project as well as the approach and scope
that will were followed towards finding a solution to this problem. A literature review explaining
the methodology and application of potential methods that were considered in the selection and
development of the final model(s) are included. Data analysis were carried out on the data that
were acquired and similarities and observations documented. Finally the appropriate methods were
tested and evaluated against one another.
The results obtained from this project indicated that it is indeed possible to develop a forecasting
tool capable of forecasting daily sales for different restaurants with similar operating
environments. It was also determined that an Expert System can be developed based on the results
obtained. The methodology that the Expert System will follow in producing forecasts are also
included in this document.
Mini Dissertation (BEng)--University of Pretoria, 2017.