The Application Of Data Mining's Decision Tree Induction In Accelerated Freight Transportation

Since the IT technology has been greatly developed, the accelerated freight transportation company's gained vast of historical data. But the shortage of analyzing tools makes it rather useless. The decision makers are still relying on their own intuition for crucial decision making. In this case, a more developed data analyzing capability in order to serve better judgment is badly needed. Data mining is a data organization and analyzing technology that based on IT. The thesis is mainly focused on data mining's effect on accelerated freight transportation's decision making application issue, including the following aspects:(1) Introduction of data mining technology, including the introduction of data mining's preparation work, general application process and common data mining technology. Also the choosing of the appropriate data mining for accelerated freight transportation: decision tree method.(2) Discussion of the decision tree's structure. And propose for a new improved algorithm.(3) Analyzing other data mining case and the data collected from CRE. All the data are after decimation and translation algorithm. Then input the "valued and clean" data into the database. Finish all the data's preprocessing in order to reduce data's relevance and also to reduce the quantity of the variable.(4) Using the improved decision trees' algorithm modeling according to CRE's data. And make the demand forecasting for freight transportation. In the modeling process, the system will delete the abnormal outlier automatically. Finally it will have a binary tree model, and the definition of this model.