Vegetable Price Forecast And Correlation Analysis Based On Time Series

Vegetable,as the largest planting area and the highest economic value crop except for food in China,is an important pillar industry for the development of agriculture and rural economy.Vegetable prices directly affect farmers’ income and residents’ consumption,and it is an important index to reflect changes of agricultural economy.In order to prevent market fluctuations caused by vegetable price changes,it is necessary to realize the accurate prediction of vegetable prices and explore the correlation and price transmission mechanism between vegetables.In this paper,the price datum of 10 kinds of vegetables in the 10 major vegetable markets from 2020 to 2021 are obtained by using python crawler,a vegetable price prediction model is established,and price correlation analysis of different kinds of vegetables are analyzed.The main contents of this thesis are as follows:1.By using a crawler program,the price datum of 10 common vegetables are obtained from 10 vegetable markets in China(Beijing Fengtai Vegetable Market,Anhui Fuyang Vegetable Market,Guangzhou Jiangnan Vegetable Market,Hefei Zhougudui Vegetable Market,Zhejiang Jinhua Vegetable Market,Nanjing Vegetable Market,Qingdao Vegetable Market,Shouguang Vegetable Market,Guangdong Shantou vegetable market,Hubei Wuhan vegetable market,Hunan Changsha vegetable market),and the MCMC algorithm is adopted to deal with outliers and missing values.2.The vegetable price prediction models of ARIMA,Prophet,and RNN deep learning are established.According to the test sets,the results show that the average RSMEs of the three models are 0.845,0.9 and 0.3,respectively.3.The vegetable price forecasting model is optimized by combining different models.According to the characteristics of data and model,the weighted average of each individual forecast is given based on the reciprocal variance method,and the RNN-Prophet forecasting model is constructed.The results show that the average RSME is 0.06,which indicates that the model effect and forecasting accuracy are significantly enhanced,and the constructed model is suitable for vegetable price forecasting.4.Find the relationship between 10 vegetables through the Apriori algorithm.The results show for Chinese cabbage,rape and cabbage,when the price of one vegetable rises,it will lead to the price rise of the other two vegetables.5.The vegetable price transmission is analyzed by using the vegetable price datum.The cointegration test results show that there is a long-term stable price equilibrium relationship between rape and Chinese cabbage.According to 5%significance level,there are two co-integration relationships,one of which indicates that Chinese cabbage,rape and cabbage are positive correlations.The 1%price change of cabbage will cause the prices of rape and Chinese cabbage change 9.1%and 20%.By using short-term dynamic relationship analysis model of the VECM,it shows that in case of short-term price mutation,the prices of rape and Chinese cabbage will reduce the deviation based on long-term equilibrium relationship,so as to promote price volatility to be equilibrium.The intensity of error correction is 20.5%.Granger causality test results illustrate that the price of Chinese cabbage is closely related to that of rap.The variance decomposition results show that the prices of rape,white cabbage and Chinese cabbage are most affected by their own residual,which indicates that these price fluctuations mainly come from their own reasons,such as bad weather,natural disasters and so on.The impulse effect analysis of price transmission efficiency shows that the price of Chinese cabbage responds quickly to the price changes of rape and cabbage.The price of rape is sensitive to the price change of Chinese cabbage for a long time.Based on the finite distribution lag model,the price transmission intensity is analyzed.The results show that the price of Chinese cabbage is mainly affected by the self price of the previous two periods,the impact degree is 19%.The price of rape is mainly affected by the price of Chinese cabbage in the current period,the impact degree is 7%.The price of Chinese cabbage is mainly affected by the price of Chinese cabbage in the current period and the previous two periods,the impact degrees are 14% and 7% respectively.