Research On Investment Estimation Model Of Municipal Road Engineering Based On Data Mining

Municipal road construction plays a crucial role in ensuring the normal operation of a city’s infrastructure.Due to the large scale and long duration of municipal road projects,they are highly influenced by external factors such as geological conditions and existing underground pipelines.Therefore,it is necessary to have reasonable investment estimation to ensure the economic benefits of the projects and the proper allocation of government funds.Municipal road projects are non-profit government investment projects,and uncontrolled investment would harm the construction of urban infrastructure.Therefore,in-depth research on investment estimation for municipal road projects is needed.Currently,the estimation of municipal road project investments faces the problem of "abundant data but lack of information".It is urgent to improve the intelligence and informatization level of investment estimation for municipal road projects and provide scientific and timely decision support.This article analyzes the current situation of investment estimation for municipal road projects and finds that the existing estimation methods are mostly based on experience and lag behind.It proposes improving the investment estimation for municipal road projects through data mining techniques and conducting research on intelligent estimation.An integrated and intelligent investment estimation model for municipal road projects is constructed.Firstly,the influencing factors of municipal road projects are identified and analyzed through the PBS project decomposition structure.Then,a set of estimation feature indicators is selected and formed based on expert survey methods.Secondly,to address the issues of scattered and missing data in municipal road projects,research is conducted on data preprocessing and data integration methods,and an investment estimation database for municipal road projects is constructed.To tackle the problem of a large number of engineering cases,a similarity case retrieval study based on the CRITIC weight method and fuzzy cluster analysis is conducted.For complex conditions in investment estimation for municipal road projects,the IABC-XGBoost-Bagging ensemble learning algorithm is used for intelligent mining and analysis,providing deterministic mean estimation results and uncertain interval estimation results.Additionally,to achieve the informatization mining of investment estimation,research is conducted on data post-processing methods such as cost composition analysis based on similar cases and feature importance analysis based on SHAP.Finally,based on the application of data mining techniques in investment estimation for municipal road projects,an architecture for the investment estimation system is designed.Through case analysis,the investment estimation model for municipal road projects constructed in this article demonstrates high accuracy and reliable interval estimation.In practical cases,compared to traditional estimation methods,the deviation between the estimated results and the actual completion settlement price using the model in this article is 3.68%,which is smaller than the 12.56% deviation of traditional methods.Lastly,this article conducts cost composition analysis and feature importance analysis for the case projects.