Highway Project Cost Estimation through Decision Tree Modeling

Highway Project Cost Estimation through Decision Tree Modeling

Manoj Chhetri1

1College of Science and Technology, Royal University of Bhutan

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract

Contractors undertake construction and maintenance projects ranging from a few months to several years, and these contracts often exhibit disparities between projected and actual costs. To illustrate, in Trashigang Dzongkhag, three out of five projects spanning 2015-2019 not only experienced cost underestimations but also encountered an average cost overrun of 27% compared to initial estimates. Furthermore, research suggests that issues in cost estimation during the project's conceptual phase can impact both project progress and expenses (Hashemi et. al., 2000). In this study, a decision tree model was employed to forecast actual costs based on initial estimations and project duration. A decision tree model with an accuracy rate of approximately 93.6% in highway project cost estimation was developed through this research. The findings indicate that the decision tree model emerged as a valuable tool for improving cost estimation accuracy.

Key Words:  Decision tree, machine learning, highway, cost estimation