Range Estimating for Risk Management Using Artificial Neural Networks

Abstract

Abstract This research developed a technique for generating range estimates to evaluate the risk of cost escalation in building construction projects, using an artificial neural network model of construction project costs. By identifying the risk of cost escalation during the planning and conceptual design of built facilities, facility owners and project managers can better focus their efforts to control total project cost in these times of decreasing capital investment budgets. The specific focus of this research was to develop a methodology for analyzing cost data from existing facilities to generate cost-probability curves and range estimates for facilities at the conceptual stage of design, that could be used in turn to identify critical variables to be managed for controlling project costs.