Analysis And Evaluation On Commodity Inventory Management Data In Circulation Enterprise

With the reduction of intermediate links in manufacturing production and marketing, more and more corporate customers are contact with inventory management directly. In order to reduce inventory costs and better meet the needs of customers, businesses need a reasonable inventory management. Inventory management is such behavior that purchase, forecasts, plans or add inventory and control of the behavior in accordance with the requirements of the outside world for stocks, the characteristics of corporate. Inventory management could manage and control the enterprise goods in the course of operation of the various finished goods, semi-finished products, raw materials, and other materials to maintain the stock at a reasonable level, to meet the production of demand at the lowest total cost of stock and stock sales.Research indicates that inventory management is focused on determining reasonable inventory levels. To determine the level of inventory related to inventory shortages and maintenance costs, enterprises must maintain a certain level of inventory so that both the total cost for the smallest. At the same time, before determining the maintenance and inventory costs out of stock, it is also necessary to determine the level of customer service level, as well as the safety stocks to maintain the predetermined level of service. Shortcomings can be found with Changes of stock level analysis and forecasts in the inventory management, so meet the needs of customers effectively. With development of information technology and data mining technology, the inventory-management-oriented data mining system will improve enterprise inventory management level.First, the paper research the status quo of inventory management and data mining technology, at present, a great number of scholars introduce the artificial intelligence to data mining of inventory management, especially the fuzzy math and neural network theory. The paper focuses on the Lobabidi fuzzy math theory and He Yan-Xiang neural network theory, both used in inventory management.Paper introduced the concept and technology of inventory management, discussed the inventory management involved in data warehousing, data mining, neural network theory and technology, on this basis discussed the specific data mining process in details.The paper analysis business processes and functions structure of inventory management system, describes the composition of inventory management system in general, including a list of materials design, keep inventory design as well as send receive design, establishment index system of the inventory management system for data mining, including inventory control costs, maintenance costs analysis, the analysis indicators of customer service level and so on. Finally, paper induced the quantitative methods of index system, to make the system of indicators that can be quantified in order to facilitate the use of computer analysis.According to the goal of data mining system of the inventory management system and business needs, paper designed the application of data warehouses, including the inventory control theme, the cost of out-of-stock theme, the customer service level theme, safety stock theme, and the structure of the four themes in data warehouse, As well as the related fact sheet and the peacekeeping table.Based on BP algorithm, the process of learning and neural network analysis model, paper describes the neural network learning process and topology model, designed the inventory data mining logic as well as the corresponding software system chart, including data sources module, data processing module, data training module, resulting output module and data maintenance module. Based on inventory management theory, established an effective indicator system of inventory analysis level, At the same time designed the model of algorithms and network analysis. The article focused on the consideration of the original BP neural network algorithm less efficient, easy local minimum, to improve the value of the right frequency of standard BP algorithm, so as to improve efficiency.Finally, the paper analyzed the real inventory data, which was from gold Lily warehouse from June 2006 to November, to verify the correctness of the design and the application feasibility of the index system, mining algorithm of data mining system. The paper has some theory and practical value.