Data Mining on Prices of Highway Project Main Material Based on WEB.pdf

Data Mining on Prices of Highway Project Main Material Based on WEB Jie He Yikai Chen Wen Hang Zhiguo Qi Transportation College, Southeast University, Nanjing 210096, P. R. China Abstract The management of main material prices of provincial highway project quota has problems of lag and blindness. Framework of provincial highway project quota data MIS and main material price data warehouse were established based on WEB firstly. Then concrete processes of provincial highway project main material prices were brought forward based on BP neural network algorithmic. After that standard BP algorithmic, additional momentum modify BP network algorithmic, self-adaptive study speed improved BP network algorithmic were compared in predicting highway project main prices. The result indicated that it is feasible to predict highway main material prices using BP NN, and using self-adaptive study speed improved BP network algorithmic is the relatively best one. Key Words: Highway project main material, Data mining, WEB, BP neural network 1. Introduction Low managing efficiency, long updating cycle and difficult in decision support are general problems in highway project management in China. For highway project management in province, these problems are especially prominent, becoming one of the most difficult problems in project fund management. These problems also have indispensable negative effects on projects’ economic and social benefits [1]-[2]. Since that, this thesis analyzed the framework of provincial highway project integrate MIS and the construction of database, data warehouse at first, based on which studied the application and implement of data mining in highway project main material prediction. Finally overcome the lag and blindness of highway project main material management, enhanced highway project finance management efficiency. 2. Data Collection and Management Based on WEB 2.1. Framework of data manag- ement syst