摘要
Abstract
The forecast and early warning of pork price can effectively guide pig breeding and ensure the supply and demand balance.In order to promote the stable and sustainable development of the pork market in Gansu province,the authors used the pork price as the research object,screened the variables by using the Lasso regression,improved the random forest by using the HHO,and the HHO-RF was established to forecast and early-warning.The results showed that the MSE is 0.22,the R2 is 99.76%,and the PRD is 20.6,which has a good fitting effect;the risk early warning effect of pork price is good,and the accuracy rate reaches 91.67%,which showes that the model HHO-RF can accurately predict the price of pork in Gansu province.The forecast for the pork price in Gansu province from November 2022 to October 2025 showed that the overall trend of prices during the forecast period is first falling,then rising and finally stabilizing.Based on this,some suggestions on the healthy development of pork industry in Gansu province were put forward from optimizing the pork price forecasting system to improve the precision of price forecasting,expanding the channel of pork price information announcement to promote the information sharing,and establishing and improving the pork price risk early warning system to improve the level of emergency regulation.关键词
猪肉价格/走势预测/Lasso回归/随机森林/哈里斯鹰优化算法/风险预警机制Key words
pork price/trend forecast/Lasso regression/Random Forest/Harris's Hawk optimization algorithm/risk warning mechanism
