Study On Inventory Demand Forecasting Of Fast Moving Consumer Goods Based On GA-BP Neural Network For A Company

As our living standard is improving,the demands of our customers are also very diversified.The competition among enterprises has became severe as they all seek to keep or even increase their share of markets.The number of retail consumer goods grew at an annual rate of more than 10%,among that,daily necessity which holds a large consumer group and continues to expand its presence in markets contributes a lot to this fast growth.Although the fast moving consumer goods industry has a bright future,internal environment of the industry is complicated and volatile,and almost all the enterprises in the industry have to experience up and down.In the course of management,enterprises should not only meet the market demands,but also improve their profitability as a way of ensuring the sustainable and stable development.It's strong alternative characteristic results the instability of the market demand.In order to guarantee the normal supply of the products,company prefers to hold a relative high level of inventory,which costs a lot and hurts their benefits.There are no technology and concept innovation recently,which means improving the accuracy of forecasting product demand almost became the only way to reduce inventory costs.Currently,related research of the fast moving consumer goods industry is mainly based on traditional quantitative prediction method,what's more,grey forecasting method,prediction of time series,regression analysis and other kinds of methods are also applied to do the research.Compared with ordinary commodities,the demand of FMCG is much more irregular for the reasons of many potential or unquantifiable factors,This paper analyzes a company's status and then forecasts the demand of its two most important products with a BP neural network model,optimizes BP network weights and thresholds with the help of introducing the genetic algorithm,finally constructs a GA-BP neural network model.In order to verify the advantages of the GA-BP neural network model in prediction of FMCG demand,this paper makes a predictive model of two kinds of products in the company with traditional methods,concerning about forecast accuracy,generalization ability and stability.According to the statistics,the GA-BP neural network model for comprehensive prediction of performance is better than any other methods.Based on the improvement of forecast accuracy,We use the EOQ model and safety stock theory to update the inventory strategy of raw materials and finished goods As is shown in the results,when we ensure a certain kind of service,there are an 49.04%decrease in original raw materials inventory cost,an 19.22%decrease of finished goods inventory cost compared with its respectively original inventory cost.This paper finally proves that the GA-BP neural network model on inventory demand forecasting for the fast moving consumer goods industry is feasible and effective.