Prediction Of The Pig Price In Shanxi Province Based On Long Short-term Memory Neural Network Model

China is the world's largest pig farming country,and the price of live pigs has fluctuated too much in recent years.The traditional forecasting method can only analyze the small data volume of the hog price,resulting in low prediction accuracy and slow speed.Moreover,the research on the hog price forecast is limited to the improvement of the forecasting method,or the short-term prediction result is not carried out.The application of reality,the people's prediction of future prices is still based on their own experience.Because the price of live pigs has many influencing factors and non-linear characteristics,the changes in the hog market also show new rules.For the pre-judgment of domestic hog prices,the past rules are not suitable,and these all put forward the prediction of hog prices a new challenge.Faced with the above challenges,this paper proposes the following solutions: Firstly,through the research on the status quo of Chinese pig production and sales,historical price data and its influencing factors,qualitative analysis of the causes and laws of hog price fluctuations.Then,based on the collection,collation and research of the historical data of pig prices in Shanxi Province for the past three years,the long-term and short-term memory neural network models and the threshold cycle unit network model are respectively carried out by means of the current advanced long-term and short-term memory neural network prediction methods.Multiple trainings have been implemented to run on its Pycharm.Comparing the two,it is concluded that the long-term and short-term memory neural network model has smaller error and better effect in forecasting,and uses it to predict the price of three kinds of pigs in Shanxi Province.It is concluded that the price of live pigs in Shanxi Province is on the rise.The price of live pigs in the next three months is a state of volatility rising,but will not rise sharply.Finally,the application of the pig market data in Shanxi Province has been implemented to realize the pig market data visualization software,which can display the price of live pigs in the previous year in various prefecture-level cities in Shanxi Province and the forecast data for the next year,providing users with more convenient and quicker pigs price access.Based on the development status of the hog industry,this paper uses advanced neural network models for forecasting,which provides a major reference for farmers and other market participants to rationally arrange procurement,production and sales plans,reducing blindness,and further enhances pig market data.Guiding significance for production and sales.