A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Predicting Budgetary Estimate of Highway Construction Projects in China Based on GRA-LASSO
Under the development strategy of traffic power, Chinese highway engineering has been in the critical stage, and controlling investment has become a top priority. The budgetary estimate is the main basis for the Chinese government to verify the highway project investment, which is estimated mainly based on the quota and price information issued by government but still lacks systematic cost indices to measure its rationality. The highway engineering cost of different regions in China varies greatly, and the accumulated cost data are not insufficient. This paper chooses the gray relation analysis-least absolute shrinkage and selection operator (GRA-LASSO) hybrid method, which has significant advantages in dealing with high-dimensional small sample data and multiple correlation variables, to establish the highway engineering budgetary estimate prediction model so as to examine the rationality of estimated investment at the initial stage of construction. Using budgetary estimate data on 121 highway projects that have been approved by the Chinese government and various macroeconomic data, the key cost drivers are identified through gray relation analysis (GRA) quantitatively, and 10 high-correlation indexes are selected as input variables. Then, the relationship among engineering characteristic variables, economic factors, and highway engineering budget are established by least absolute shrinkage and selection operator (LASSO) regression. The empirical test results show that eight indexes have significant impact on the highway cost, including earthwork quantities, route length, number of lanes, bridge-tunnel ratio, protection works quantities, highway construction investment, producer price index (PPI) and purchasing manager’s index (PMI) of construction industry. Compared with ordinary least-squares (OLS) regression and LASSO method alone, the GRA-LASSO combined method has better performance in error evaluation index and stronger model accuracy and generalization ability. The study result can be mainly used for government to rapidly audit, forecast, forewarn, monitor, and manage the highway engineering cost, and also help investors and designers to pay close attention to the influence of design parameters and external environment on cost. Moreover, it provides a reference method or model for establishing a highway engineering cost index.
Predicting Budgetary Estimate of Highway Construction Projects in China Based on GRA-LASSO
Under the development strategy of traffic power, Chinese highway engineering has been in the critical stage, and controlling investment has become a top priority. The budgetary estimate is the main basis for the Chinese government to verify the highway project investment, which is estimated mainly based on the quota and price information issued by government but still lacks systematic cost indices to measure its rationality. The highway engineering cost of different regions in China varies greatly, and the accumulated cost data are not insufficient. This paper chooses the gray relation analysis-least absolute shrinkage and selection operator (GRA-LASSO) hybrid method, which has significant advantages in dealing with high-dimensional small sample data and multiple correlation variables, to establish the highway engineering budgetary estimate prediction model so as to examine the rationality of estimated investment at the initial stage of construction. Using budgetary estimate data on 121 highway projects that have been approved by the Chinese government and various macroeconomic data, the key cost drivers are identified through gray relation analysis (GRA) quantitatively, and 10 high-correlation indexes are selected as input variables. Then, the relationship among engineering characteristic variables, economic factors, and highway engineering budget are established by least absolute shrinkage and selection operator (LASSO) regression. The empirical test results show that eight indexes have significant impact on the highway cost, including earthwork quantities, route length, number of lanes, bridge-tunnel ratio, protection works quantities, highway construction investment, producer price index (PPI) and purchasing manager’s index (PMI) of construction industry. Compared with ordinary least-squares (OLS) regression and LASSO method alone, the GRA-LASSO combined method has better performance in error evaluation index and stronger model accuracy and generalization ability. The study result can be mainly used for government to rapidly audit, forecast, forewarn, monitor, and manage the highway engineering cost, and also help investors and designers to pay close attention to the influence of design parameters and external environment on cost. Moreover, it provides a reference method or model for establishing a highway engineering cost index.
Predicting Budgetary Estimate of Highway Construction Projects in China Based on GRA-LASSO
Tong, Bin (author) / Guo, Jingjuan (author) / Fang, Shen (author)
2021-02-26
Article (Journal)
Electronic Resource
Unknown
Forecasting Completed Cost of Highway Construction Projects Using LASSO Regularized Regression
| Online Contents | 2017
Subsequent assessment about China highway construction projects
| British Library Conference Proceedings | 1998
Modeling the Construction Risk Ratings to Estimate the Contingency in Highway Projects
| Online Contents | 2017