Hog Price Analysis And Prediction Based On The BP Neaural Network

In recent years,the hog price is fluctuating frequently in China,which have not only caused huge economic losses to the producers and operators,but also causes great troubles to the consumers.And it's not conductive to the healthy and sustainable development of the hog market,which in turn adversely affects the argricultural product market.How to accurately grasp the fluctuation rules and cycle of hog price and accurately predict the hog price is particularly important.In this paper,Firstly,using the weekly data of China's hog price to research and analyze the hog price,and draw the basic annual and inter-annual fluctuation rules and cycles of hog price in China.Based on this analysis,the factors affecting the price fluctuation of hog price in China are analyzed.Mainly from the supply factors,demand factors and external factors in three aspects of a comprehensive analysis.Then,using grey correlation analysis method,correlation coefficient analysis method,qualitative analysis method and Stepwise regression analysis method to analyze the influence factors that influence the hog price,and analyze the significant influence factors that influence the hog price fluctuation.Finally,using BP neural network prediction model and multiple regression prediction model,a hog price prediction model based on BP neural network and multivariate regression analysis is established,namely BP-multiple regression prediction model,which comprehensively consider the impact of historical hog prices and influencing factors on the hog price predicton.According to empirical analysis,during the year,hog price exhibited the characteristics of “two highs and low in the middle”,during the year hog price have a three-year volatility cycle,pork price,piglet price,corn piece and soybean meal price have a significant impact on the fluctuation of hog price,and the prediction accuracy of BP-multiple regression prediction model is as high as 96.15%,which is more than 11% higher than the prediction accuracy of a single BP neural prediction model and multiple regression prediction model.Therefore,the BP-multiple regression prediction model has a better prediction effect on the hog prediction.