Prediction Of Pig Price Based On Time Series

China is a large agricultural country with a long history and pig farming is an important part of it.Since ancient times,China has had a saying that "pig food secures the world",which highlights the important role of the pig industry in production and life.However,in recent years,pig prices have fluctuated abnormally.Frequent fluctuations in the pig market can reduce the motivation of farmers and affect the daily lives of consumers,leading to abnormal changes in the overall CPI,which in turn can have a negative impact on economic stability.Therefore,it is particularly important to analyze the characteristics of pig price fluctuations in China and to predict pig prices.This study aims to analyze the volatility characteristics of pig prices and the factors influencing them in China as a research object,build different forecasting models based on them and finally develop a visualization system to present the results.The main research includes the following three aspects:1.Analysis of pig price characteristics and influencing factors.The seasonal adjustment method and H-P filter method were used to decomposition analysis of pig prices in China to understand their trend,cyclical,seasonal and stochastic characteristics.Using price fluctuation and grey correlation theory,the main factors affecting the fluctuation of pig prices are analyzed based on the relationship between supply and demand and other relationships of hogs.2.Model construction.In view of the seasonal characteristics of pig prices,this paper proposes to use the SARIMA model which considers seasonal factors;in view of the general performance of traditional neural network models in the field of pig price forecasting,this paper proposes to incorporate the attention mechanism Attention into the LSTM model and compare it with other models for analysis.3.System development and prediction.Based on the three hog price prediction models established,the system is developed using Spring Boot to encapsulate the prediction models and provide the front-end operating interface and prediction results viewing functions,which can effectively reduce the tediousness of obtaining prediction results.In conclusion,based on the analysis of the characteristics and influencing factors of pig price fluctuations,this paper establishes a SARIMA model considering the seasonal characteristics of pig prices,and incorporates the attention mechanism Attention into the LSTM model in order to improve the accuracy of the prediction model,and establishes multiple forecasting models for comparative analysis.Finally,based on the three forecasting models established,Spring Boot was used to develop the system and encapsulate the forecasting models so that the front-end users can operate and view the forecasting results directly,realizing the visual presentation and use of the models.