| Title |
A Study on the Assessment Model of Preliminary Cost Estimates Using Support Vector Machines |
| Authors |
Park U-Yeol ; An Sung-Hoon ; Kang Kyung-In |
| Keywords |
Assessment Model ; Preliminary Cost Estimates ; Support Vector Machines ; Accuracy |
| Abstract |
Clients have wanted to assess the quality of preliminary cost estimates, because construction costs are estimated under a high level of uncertainty about the project future. Therefore, the object of this study was to develop the assessment model of preliminary cost estimates using support vector machines(SVM). The factors that influenced the preliminary cost estimates are selected, and 58 projects, which were actually estimated and constructed, were collected from the construction companies. SVM was used to develop the assessment model of preliminary cost estimates. In order to develop the SVM model, the binary SVM classifier is expanded into a multi-class classifier. and to present the feasibility of our SVM model, we considered 58 projects. The results showed that our SVM model classify the accuracy ranges of preliminary cost estimates efficiently. In addition, the selected key factors, which highly influenced the quality of cost estimates, will be used to improve the accuracy of the preliminary cost estimates. |