A CBR-based hybrid model for predicting a construction duration and cost based on project characteristics in multi-family housing projects

作者：Koo ChoongWan; Hong TaeHoon^*; Hyun ChangTaek; Koo KyoJin

来源：Canadian Journal of Civil Engineering, 2010, 37(5): 739-752.

DOI：10.1139/L10-007

• construction costs
• housing
• regression analysis
• Monte Carlo method
• neural networks
• optimization

Decision-making in the early stage of a project has a significant impact on the project. However, limited and uncertain information on the project and a complex correlation among various factors that affect the project's construction duration and cost, make it difficult to predict and manage the project. Therefore, this study developed a case-based reasoning (CBR)-based hybrid model with which to predict the construction duration and cost of a project in its early stage. One hundred and one cases among multi-family housing projects that were completed between 2000 and 2005 were used. The CBR-based hybrid model developed in this study is the result of integrating the advantages of (i) prediction methodologies, such as case-based reasoning, multiple regression analysis, and artificial neural networks, (ii) the optimization process using a genetic algorithm, and (iii) the probability distribution and the analysis process of outlier using Monte-Carlo simulation. The results of this study are expected to support the owners and managers who are in charge of estimating budget and construction duration in both public and private sectors, in predicting accurately the construction duration and cost at the business planning or early stage of a project.

• 出版日期2010-5