摘要
我国是畜产品生产消费的大国。随着畜产品种类的日益丰富,畜产品消费结构也悄然变化,为了避免市场供需失衡,研究预测我国畜产品消费量对引导制定较为合理的生产计划有着十分重要的意义。基于此,构建了基于灰色关联分析和支持向量回归机的畜产品消费量组合预测模型。灰色关联分析为组合预测提供了选取单项预测模型的依据,确保了参与组合预测的单项预测模型的质量,支持向量回归机以其良好的学习泛化能力用于组合预测中,可以对复杂环境下的事物做出较为准确的预测。在上述理论与方法研究的基础上,将基于灰色关联分析和支持向量回归机的组合预测模型综合应用于我国猪肉消费量的预测实践中,通过实验结果比较分析,验证了研究成果的有效性。
China is a big country in the production and consumption of livestock products.Because the variety of livestock products is increasing,the consumption structure of livestock products has also been changed.In order to avoid imbalance between supply and demand in the market,it is significant to study and predict the consumption of livestock products in China,which can help to guide the formulation of more reasonable production plans.Thus,a combinational forecast model based on Grey Correlation Analysis and Support Vector Regression is built.Grey Correlation Analysis provides the guides for selecting individual forecast model for combinational forecast model to ensure the quality of forecast.The Support Vector Regression is used in combinational forecast because of its good generalization ability,which can be used in complex environments.According to the above theoretical research,the combinational forecast model based on Grey Correlation Analysis and Support Vector Regression is applied to the prediction of Chinese livestock product consumption.The effectiveness of the research is verified by comparative analysis.
作者
吉敏
李卓
JI Min;LI Zhuo(School of Intelligent Manufacturing and Control Engineering,Shanghai Polytechnic University, Shanghai 201209,China;School of Agriculture and Biology;Institute of New Rural Development, Shanghai Jiao Tong University,Shanghai 200240,China)
出处
《上海第二工业大学学报》
2018年第4期285-291,共7页
Journal of Shanghai Polytechnic University
基金
上海第二工业大学校基金项目(EGD17XQD12)资助
关键词
灰色关联分析
支持向量回归机
组合预测
单项预测模型遴选
畜产品消费量
grey correlation analysis
support vector regression
combinational forecast
individual forecast model selection
livestock product consumption
作者简介
通信作者:吉敏(1987-),女,上海人,讲师,博士,主要研究方向为企业信息化。E-mail:jimin@sspu.edu.cn.
