An expert system for construction cost indexing and forecasting

摘要

Project planners need to forecast the cost of construction for a future project in order to determine its feasibility. There are numerous approaches to performing such a task. This paper presents some of these approaches, and extends their validity by correlating, not only the inflation adjusted construction cost indices, but by generating a geographical correlation between indexed cities. The paper also provides a statistical confidence factor that gives an indication of the percentage prediction error based on the level of correlation between one city's historical price curve with another city's historical price curve. Two approaches are presented. The first approach is an inflation adjusted historical comparison of a project cost to another known project cost. The second approach is a regression analysis that compares project costs at different locations, and different time periods.