Decision making model for selecting subcontractor relationship in Cambodia construction project

Abstract

The objective of this research is to develop a model that could help main contractor to evaluate subcontractor on selecting short or long-term relationship in Cambodia construction projects. There are 2 stages that need to accomplish in this model development. First, it identified the important factors for selecting subcontractor relationship by interviewing with 35 main contractors who are directors or project managers. From the analysis result, there were ten of twenty two factors that main contractor considered as the important factors for selecting subcontractor relationship namely time control in planning, work quality, cooperation, experience, commitment, resources, honesty, trust, monitoring, and coordination. Second, the model was developed by using Artificial Neural Network (ANN) and Discriminant analysis. Then, the models were developed and selected a suitable method that has the high level of accuracy. The selection of method was based on 10 factors with 93 samplings in which main contractor evaluated subcontractors. The result of ANN model shows the low percentage of root mean squared error (RMSE) around 0.02 and 0.04 for training and testing data set. In discriminant analysis, the percentages of accuracy in estimation are 98.7 % whereas the validations are 83.75 %. Thus, our research takes the neural network as an optimal method for developing a model because this method has given a good result of low error. Finally, this research would be useful in main contractor decision making on selecting subcontractor for developing a long-term relationship.