Dynamic Demand Forecasting Of Fresh Agricultural Products In E-Commerce Circumstance

With the rapid development of socio-economic, urban residents’ pace of life is accelerating gradually. People in the pursuit of high-quality fresh agricultural products are also seeking convenient way to buy fresh agricultural products meanwhile. The emergence and development of fresh agricultural products e-commerce meet people’s demand of buying high-quality fresh agricultural products through convenient ways. It also builds new sales channels between the producers and consumers of agricultural products.In the age of Big Data, the value of information to the countries, enterprises and individuals is no longer limited to the way of obtaining. Instead it reflects more in the way how to extract and analyze the data effectively with the known information in order to guide the development of national policies, improve the corporate strategy and make individuals’ lives better. The goal of fresh agricultural products e-commerce is to solve the problem of supply and demand in agricultural markets completely and make effective coordination of supply and demand balance between the producers and consumers of agricultural products in order to make agricultural production in order form. To achieve this goal can’t be separated for the processing of data in the information of fresh agricultural products e-commerce. Only with processing and analyzing the data effectively, the market trends can be grasped and the right balance between supply and demand can be found. Therefore, the accurate and in real-time forecasting of fresh agricultural products market information becomes increasingly important.The article first summarizes some existing domestic fresh agricultural products e-commerce model, investigates some shortcomings restricting its development, and proposes a new e-commerce with the cooperation of e-commerce enterprises, high level agricultural production bases, third-party cold chain logistics enterprises and community stores. With the analysis, it also pointed out that dynamic demand forecasting is the key factor in the impact of fresh agricultural products e-commerce development. The article based on a variety of factors that influence changes in demands, combined with fuzzy quantification data preprocessing techniques, uses RBF neural network and SVM (support vector machine) to construct a dynamic demand forecasting system of fresh agricultural products. Then a real case is analyzed under the forecasting system using two methods with the application of MATLAB simulation software in order to compare their effectiveness. Finally, the main conclusions and shortcomings are summarized, and make prospect of the research in related fields.The results show that the dynamic demand forecasting system using nonlinear regression model under multi-factor structure, can forecast the demand of agricultural products in e-commerce circumstance in real-time and effectively. It makes a positive contribution to solving the problem of unbalanced supply and demand.