Acquisition Price Forecast Of Yantai Apple Based On BP Neural Network

ABSTRACT: Yantai is a big city in our country which exporting apples. Yantai apple’s grown and cultivation is always a pillar industry of Yantai economy. In recent years the market price is often fluctuation and amplitude are also great, Yantai varieties often have to face the awkward situation of "fertility not harvest". In reaction to the phenomenon, this paper has a deep investigation, and then select main factors may affect the acquisition price of Yantai apple, hopes to use the data the BP neural network’s function of predicting the future observations to establishment of market prediction model of the purchase price. With the help of this model, Yantai growers can cope with market change better and due to economic losses for lacking information or deciding production blindly.This paper investigated and collected the data of Yantai apple’s market purchase price from2004to2011as the dependent variable, trying to using data analysis to predict purchase price trends in the future. According to the equilibrium price theory, this paper established analytical framework of affecting factors from aspects of demanding and supplying, and selected nine main factors: production, inventories, factor prices, revenue for selling apples, distribution cost (highway mileage and automotive fuel price index), Urban disposable income, Alternative products (Shanxi apple and Zhejiang citrus) prices, In order to ensure that the selection of influence factors accurately reflect the Yantai apple purchase price and ensure the neural network of the evaluation results’scientific and objectivity, this paper does simple correlation analysis firstly.Compared to other models, the neural network’s predicting error is small. This model has advantages of a simpler theory, lesser data calculation, stronger parallel sex, actual simpler operation, etc. Based on the correlation analysis, we choose six factors as the neural network’s input variables. Through exploring step by step, such as the network layer, hidden neurons number, initialized weights, learning rate and expected error, finally this paper establishes Yantai apple’s purchase price prediction model. At last, this paper use the model to forecast the price of2012, and the results prove that price will continue to go up.