基于梯度提升回归树模型的烟草产量预测方法  

Tobacco Output Prediction Method Based on Gradient Boost Regression Tree Model

作　　者：李明钊 李熠胥 王佳 Li Mingzhao;Li Yixu;Wang Jia(Hongyun Honghe Tobacco(Group)Co.,Ltd.Kunming Cigarette Factory,Kunming,Yunnan 650106,China;Department of Automation,Kunming University of Technology,Kunming,Yunnan 650106,China)

机构地区：[1]红云红河烟草(集团)有限责任公司昆明卷烟厂,云南昆明650106 [2]昆明理工大学自动化系,云南昆明650500

出　　处：《云南化工》2023年第9期109-111,共3页Yunnan Chemical Technology

摘　　要：烟草作为我国重要经济作物之一,其利税为国家和地方财政收入作出了积极贡献。基于我国近年来烟草产量的历史数据,建立梯度提升回归树模型,对烟草产量进行了预测。首先,根据梯度提升思想建立梯度提升回归树模型;然后,根据烟草产量与年份、月份及上年同期产量间的关联,设置独立因子;最后,借助2017~2021年全国烟草产量的真实数据,对2022年同期产量进行预测分析,并与2022年全国烟草产量的真实数值比较,以验证梯度回归树模型预测的有效性。Tobacco is one of the most important economic crops in China,and its tax revenue contributes positively to national and local fiscal revenue.In this paper,based on the historical data of tobacco production in China in recent years,we establish a gradient boost regression tree model to forecast tobacco production.Firstly,a gradient boost regression tree model is established based on the idea of gradient boosting.Then,we set independent factors based on the correlation between tobacco yield and year,month and the same period of the previous year.Finally,the real data of national tobacco production from 2017 to 2021 were used to predict the production in the same period of 2022 and compared with the real values of national tobacco production in 2022 to verify the validity of the gradient boost regression tree model prediction.

关 键 词：梯度提升回归树 烟草 产量预测 

分 类 号：S117[农业科学—农业基础科学]