Predicting Profit Performance of International Construction Projects

Abstract

This study developed a practical case-based reasoning (CBR) model to predict the profit performance of international construction project (ICP) due to the nature of the prediction as a project-based, unstructured, and small sample issue. The proposed model helps contractors to make reasonable bid decision or risk treatment in the early stage of ICPs by retrieving similar cases. Grey relational analysis is applied to respectively measure the weight and variable similarity so that case similarity can be scored. Finally, the proposed model is tested on the basis of 30 actual overseas rail projects. The results indicate that the proposed CBR model performs at a higher level of accuracy than other classical CBR models. A structured process to choose appropriate methods for CBR model development by considering the prediction accuracy performance of methods available is also included.