Due to market competition, construction companies often place low bids when tenders are invited for domestic public construction projects. Over-competition can lead to vicious price wars to win a tender, which can in turn seriously affect the quality of construction. This study aims to establish an accurate Taiwan based model for the forecasting of the tendered price for roadway construction. This model is designed to assist the public sector to determine what would be a reasonable reserve price or award price. In order to ensure accurate predictions, a data classification system is established using fuzzy set theory. For each category of classified data, multiple regression analysis is applied to the linear model, the power series model, and the refined power series model. Multiple factors in the regression for the tender price prediction include the contract schedule, the budget price, and the tender bond. It is shown that the average relative error of the final reserve price model is about 3%, while that for the price of award model is 9%. In comparison, the developed reserve price model is more feasible than the price of award model.
Lin, Jeng-Wen; Chen, Cheng-Wu; and Hsu, Ting-Chang
"FUZZY STATISTICAL REFINEMENT FOR THE FORECASTING OF TENDERS FOR ROADWAY CONSTRUCTION,"
Journal of Marine Science and Technology: Vol. 20:
4, Article 9.
Available at: https://jmstt.ntou.edu.tw/journal/vol20/iss4/9