Forecasting Engineering Costs: Two Case Studies

Abstract

In recent years the subject of statistical forecasting has been brought to a sharp focus as a major area of engineering management decision-making process by the growth of substantive literature and by the onset of sophisticated computers. The choice of the most suitable forecasting technique for a particular engineering application is by no means a simple decision. This study presents two statistical methods that fully make use of the available historical data — the Box-Jenkins stochastic method and the Holt-Winters smoothing method. The forecasts made by both the methods agree very well with the actual observations.