Font Size: a A A

Based On Data Mining Business Information Systems Design And Implementation,

Posted on:2009-07-11Degree:MasterType:Thesis
Country:ChinaCandidate:H ChenFull Text:PDF
GTID:2208360245967357Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
With the development of market economy and the globalization of trade, enterprises are facing increasingly fierce competition in the market. Customs import and export bills of lading from the surface are trade data, but in the actual combat of international trade, one for each trade is very far-reaching significance to enterprises, It is a challenging problem to how used massive bill of lading data information to guide enterprises to make the right strategic decision. Based on the study of data mining and business intelligence on the basis of data mining technology ,the paper apply data mining technology to analysis of the import and export bills of lading data, which provide a new technology for make the scientific strategic decision.Firstly, this paper analyze the current business intelligence and the status quo of data mining development, then make use technology of data warehouse to establish a data warehouse of business information system, which preprocess to the customs bill of lading data. It is a premise to further provide the basis for data mining.Then, the paper analyze detailed the BP neural network technology. It study the principle of BP neural network forecasting demand for the commodity .By introducing learning momentum factor and variable rate, BP neural network not only guarantee the stability of network training, but also greatly enhance BP neural network learning speed.Finally, the papers also discuss technology of data mining for association rules. For the original association rules incremental update algorithm can not find the development trend of the rules, a rule can be found that the development trend of incremental rules mining algorithm, which can be found not only the rules of standard increment association rules, but also found a "conspicuous performance " rules.
Keywords/Search Tags:Data warehouse, Data mining, BP neural networks, Association Rules
PDF Full Text Request
Related items