Design & Study On Forecasting System Of Shipping Market Based On Data Mining

This paper is related to the key project of COSCO—shipping enterprise decision supporting system. It mainly focuses on forecasting system of shipping market, aims to provide the important economic parameter for the decision-maker and forecast the relevant factor of future development trend, thus availably holds the development trend of the shipping market;it is very important for improving the enterprise of adaptive ability and competitive ability in the international market and promoting the enterprise of management efficiency and economic performance.Firstly, in this paper the fundamental of data mining and its application are introduced, and the theory and application of Rough Sets and Adaptive Neural Fuzzy Inference System which are two important methods of data mining is summarized. Secondly, the algorithm that Rough Sets is combined with ANFIS and its application process are discussed, and then this algorithm is applied into shipping market. The error curve and forecast result show that this method is feasible. Finally, according to the shipping market forecast contents and enterprise need, the total project and architecture of this system are designed, the design of background database and specific realization of each function module are discussed in detail, and some key technique of application development are given.The forecast system of shipping market based on data mining mainly realizes addition, modification, deletion, inquiry and forecast of the related information in shipping market. In the module of shipping market forecast, breaks through the traditional method of shipping market forecast, introduces data mining to this research field, applies the combination algorithm of Rough Sets and ANFIS to establish the shipping market of forecast model that is able to forecast the demand, supply and BDI(Baltic Dry Index) of shipping market.