【摘要】 准确预测、预警猪肉价格对稳定农产品市场意义重大。本文采用LASSO回归-梯度提升回归树（GBRT）组合模型对猪肉价格进行预测，再利用黑色价格预警方法对猪肉价格预警机制进行实证研究。结果表明，玉米价格、豆粕价格、小麦麸价格、育肥猪配合饲料价格、活猪价格、能繁母猪存栏量、规模以上定点屠宰企业生猪屠宰量等指标是2009年1月至2023年12月的研究期间内影响猪肉价格波动预测预警的主要因素，据此构建的GBRT模型能够较好拟合猪肉真实价格变动趋势；运用黑色预警方法构建的猪肉价格预测警度和实际警度基本吻合。由此提出完善价格预测预警机制，增强风险识别；完善储备体系，精准引导产能调整；完善政策支持体系，夯实稳产保供能力等政策启示，以保障生猪产业链供应链安全稳定运行。更多还原

【Abstract】 Accurate prediction and early warning of pork prices are of great significance to stabilizing the agricultural product market. This paper uses the lasso regression gradient boosting regression tree combination model to predict the pork price, and then uses the black price early warning method to empirically study the pork price early warning mechanism. The results showed that corn price, soybean meal price, wheat bran price, fattening pig formula feed price, hog price, breeding sow stock,pig slaughter volume of designated slaughtering enterprises above designated Size and other indicators were the main factors affecting the prediction and early warning of pork price fluctuations during the research period from January 2009 to December 2023. The GBRT model constructed based on this can better fit the real price change trend of pork. The prediction warning degree of pork price based on the black warning method is basically consistent with the actual warning degree. Based on this, it is proposed to improve the price forecasting and early warning mechanism, enhance risk identification, improve the reserve system to accurately guide the adjustment of production capacity, improve the policy support system and consolidate the ability to stabilize production and supply, so as to ensure the safe and stable operation of the supply chain of the hog industry chain.更多还原