Research On The Short-term Demand Forecasting Model Of Meat Products In Small And Medium Sized Enterprises

In recent years, with the promotion of China's international influence, the competitive pressures faced by domestic food enterprises are increasing. Small and medium enterprises must promote the reform of their management system, in order to actively participate in international competition and quickly respond to changes in the domestic industrial structure. At present, China's food enterprises generally face the problem of excessive inventory, which has become a bottleneck restricting the development of enterprises, and the characteristics of meat products, perishable and high-cost-storage, makes the problem particularly prominent in meat food enterprises. The short-term demand forecast inaccuracy is the main reason for this problem, so the study on the short-term demand forecast of meat products for small and medium sized enterprises is necessary.In view of this, the short-term demand forecasting model of meat products is constructed based on grey theory, BP (Back Propagation)neural network, SVM (Support Vector Machine) and GA (Genetic Algorithm). The prediction model with higher accuracy and better goodness of fit is selected through experimental comparison.Firstly, the main influencing factors of short-term demand for meat in small and medium enterprises are selected. According to the selection principle of influencing factors, the 5 main influencing factors of monthly consumer price index, product price, promotion cost, seasonal coefficient and holiday coefficient are selected. Secondly, the prediction methods are selected. The artificial intelligence prediction methods with high prediction accuracy are selected based on the analysis of the existing prediction methods and the comparison of traditional forecasting methods and artificial intelligence forecasting methods. And the BP neural network and SVM are selected from artificial intelligence. Then,Prediction models are constructed. Grey prediction theory is combined with BP neural network and SVM, so that the problem that the prediction accuracy is reduced because of the randomness of the sample data is avoided. And GA is introduced to optimize the two prediction methods,in order to solve the problem that the initial parameters of BP neural network and SVM are random. Finally, the experiment is carried out and the results are compared. The grey BP neural network, grey GA-BP neural network, grey SVM and grey GA-SVM forecasting model are used to forecast the monthly demand of meat products, and the four models are evaluated based on the posterior error test method and the residual test method. The results show that the grey GA-SVM model has a higher prediction accuracy and a higher degree of goodness of fit for the small and medium sized enterprises.