Study and Realization of a Forecasting System for Steel Selling Based on Bayesian Dynamic Model
Shen Dong · Jisuanji fangzhen · 2008
According to steel sale of a large steel company,the paper has built a database by using oracle database based on vast amounts of steel selling data of the steel company,and proposed a forecasting analysis model by means of DLM(Dynamic Linear Models),then the forecasting analysis system used for data mining has come into being finally.The on-site system running has verified the correctness of the constructed model through comparing real value with predicted value,and the model also reaches a higher forecasting precision.The predicted value has important reference value for managers to make decisions.The system has been tested and proven to be useful in practice,and also can be promoted in the field of steel industry.