| 生态足迹及其影响因子的偏最小二乘回归模型与应用 |
| |
| 引用本文： | 吴开亚,王玲杰.生态足迹及其影响因子的偏最小二乘回归模型与应用[J].资源科学,2006,28(6):182-188. |
| |
| 作者姓名： | 吴开亚 王玲杰 |
| |
| 作者单位： | 厦门大学,经济学院,厦门,361005 |
| |
| 基金项目： | 教育部人文社会科学规划项目;中国博士后科学基金 |
| |
| 摘 要： | 生态足迹作为一种非货币化定量测度可持续发展的有效方法，自提出以来在国内外获得广泛的应用和发展，但有关生态足迹的时序分析以及生态足迹与其影响因子的动态研究还处于探索阶段。本文利用自然资源生产与消费数据、人口数据，计算了安徽省1990年～2003年的生态足迹，表明研究期内人均生态足迹由1.0121hm2升至1.7659hm2，增长率为74.48%。选取GDP、固定资产投资等若干重要的生态足迹影响因子作为自变量，探讨生态足迹与其影响因子的偏最小二乘回归模型的建立步骤和方法，并以安徽省近14年的相关数据进行实例分析，得出各影响因子对生态足迹的相关影响程度依次为人口、GDP、固定资产投资和居民消费支出。比较偏最小二乘回归与普通最小二乘回归进行生态足迹及其影响因子相关性分析的结果差异，揭示了偏最小二乘回归能解决多元回归分析中变量的多重相关性问题，可提高相关分析的精度，更符合实际情况。
|
| 关 键 词： | 偏最小二乘法 生态足迹 影响因子 安徽省 |
| 文章编号： | 1007-7588（2006）06-0182-07 |
| 收稿时间： | 2006-02-24 |
| 修稿时间： | 6/1/2006 12:00:00 AM |
Partial Least Square Regression Model of Ecological Footprint and Its Influencing Factors |
| |
| WU Kai-ya and WANG Ling-jie.Partial Least Square Regression Model of Ecological Footprint and Its Influencing Factors[J].Resources Science,2006,28(6):182-188. |
| |
| Authors: | WU Kai-ya and WANG Ling-jie |
| |
| Abstract: | As an effective non-monetary method to measure the sustainable development in quantitative way, ecological footprint has been widely applied and developed since it was raised. However, the time series analysis of ecological footprint is only in the initial stage domestically, so is the dynamic analysis of the relationship between ecological footprint and its influencing factors. This paper selects several important influencing factors as independent variables such as gross domestic products and total fixed assets investment to discuss how to build a partial least square regression model and how to design the steps of the model between ecological footprint and its influencing factors. The time series change of Anhui province's ecological footprint from 1990 to 2003 was worked out by using the data of production and consumption of natural resources and population. The result shows that ecological footprint per capita increased from 1.0121hm2 to 1.7659 hm2 with the growth rate of 74.48% in the research period, but less than ecological capacity and the ecological deficit has been rising year by year. Based on the data of population, gross domestic products, primary industry products, secondary industry products, tertiary industry products, total fixed assets investment, non-agricultural households consumption and agricultural households consumption, and the results of Anhui province's ecological footprint in these 14 years, the multiple regression model is built by the partial least square regression method to study the correlation of various influencing factors with ecological footprint. It is found that ranking of impacts of independent variables to ecological footprint are population, gross domestic products, total fixed assets investment and households consumption. The differences between the results of the partial least square regression and ordinary least square regression model have been compared based on analyzing the correlation of ecological footprint and its influencing factors. Due to the strong correlation among these factors, the result of ordinary least square regression model shows negative correlation between these factors and ecological footprint. On the other hand, these problems can be conquered by the partial least square regression model, including multiple correlation problems between ecological footprint and its influencing factors, even influencing factors themselves as well, which can increase the precision of correlation analysis. In other words, this model is accurate and close to the reality more than the previous ordinary least square regression model. |
| |
| Keywords: | Land use Landscape Change matrix Markov model |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《资源科学》浏览原始摘要信息 |
| 点击此处可从《资源科学》下载免费的PDF全文 |
|
