This paper uses Data Mining and Data warehouse technology to resolve thedeficiencies of traditional Data warehouse management information system, and to realizethe follow purposes: 1,filtering of large data; 2,reasonably arrangement of productionplans; 3,handling of emergency; 4,adjustment of the same stuff of differ customers;5,removing the customers of costhigh and profitlow.This paper analyses the SMT database, then builds the data warehouse logical modelwith the method of starschema, after that, The various production data, cost etc and thefluctuation condition of material, plan are loaded into the data warehouse, and to analysisthem by the primary component analysis method. One model is built by the improveddynamic cluster method. The quality of the model's result will be improved with deletedthe outlier data of cost high, fluctuation condition big etc. At last the production decisionmodel is gotten by statistics method. |