基于随机森林方法的碳足迹及其影响因素研究  

Study on the Impact Factors of the Carbon Footprint Based on Random Forest

作　　者：唐建荣 陈实 

机构地区：[1]江南大学商学院,江苏无锡214122

出　　处：《统计与信息论坛》2013年第10期70-74,共5页Journal of Statistics and Information

基　　金：国家社会科学基金项目<实物期权条件下的公共资源交易问题研究>(08BYJ060);教育部人文社科基金项目<区域性碳交易平台的系统研究>(11YJAZH084);江苏省研究生培养创新工程项目<基于多目标约束的碳减排路径及政策研究>(CXLX11_0508)

摘　　要：研究碳足迹持续增长动因对碳足迹的抑制及其碳减排政策的制定具有深远意义。基于随机森林方法并利用1997—2011年的相关数据构建了碳足迹影响因素分析模型。模型对碳足迹影响因素进行了有效识别和系统归集,并为碳足迹的控制提供了理论依据和数据支撑。研究表明:产业结构、固定资产投资、发展规模、人口规模、住宅建筑、技术水平和城市化水平等因素是形成碳足迹的稳定力量;能源结构、经济速度和碳汇效应的影响并不显著。The rapid development of economy has obviously promoted the growth of the carbon footprint, and the result makes the ecosystem which we depend on for survival more fragile. To identify and reveal the factors of the continued growth of the carbon footprint is important to make carbon reduction policies Based on the method of random forest, this paper uses the relevant data of 1997-2011 and then constructs the model that analyses the factors of carbon footprint. The model effectively identifies and systematically summarizes the carbon footprint influencing factors,and provides a theoretical basis and data support for the control of carbon footprint. The results show that： the industrial structure, investment in fixed assets, the scale of development, the population, residential construction, technical level and the level of urbanization and other factors is a stabilizing force for the formation of carbon footprint; energy structure, rapid economic growth and the impact of the carbon sink effect is not significant.

关 键 词：能源碳足迹 随机森林方法 偏最小二乘法 

分 类 号：X22[环境科学与工程—环境科学]