| 基于偏最小二乘回归的城市生活用水量预测研究 |
| |
| 引用本文： | 张立杰.基于偏最小二乘回归的城市生活用水量预测研究[J].科技通报,2012,28(2):179-181. |
| |
| 作者姓名： | 张立杰 |
| |
| 作者单位： | 梧州学院,广西梧州,534002 |
| |
| 基金项目： | 广西教育厅科研项目(201012MS202);梧州学院院级重点项目(2010B009) |
| |
| 摘 要： | 以城市生活用水量为预测研究对象,选取6个社会经济发展因素作为主要变量因子,建立偏最小二乘回归模型。研究分析表明,各变量因子间存在较强的多重共线性,采用偏最小二乘回归模型能有效克服各类因子变量间的多重共线性对模型拟合精度及其预测能力的影响,取得更接近现实的预估结果(平均相对误差为2.7%)。研究还发现,数据序列的长度和变量近期的变化信息也会对模型的预测精度产生重要的影响。
|
| 关 键 词： | 城市生活用水量 多重共线 偏最小二乘回归模型 |
Estimation of Urban Water Demand Using Partial Least Squares Regression Model |
| |
| ZHANG Lijie.Estimation of Urban Water Demand Using Partial Least Squares Regression Model[J].Bulletin of Science and Technology,2012,28(2):179-181. |
| |
| Authors: | ZHANG Lijie |
| |
| Institution: | ZHANG Lijie(Wuzhou University,Wuzhou 543002,China) |
| |
| Abstract: | the urban water demand is estimated by the least squares regression models which considers six social and economic factors as model input variables.The results show:There is a high multicollinearity among the input variables,the partial least squares regression model can reduce the undesirable effects of multicollinearity,and get a more accurate predict(the relative error is only 2.7%).It has also been found that the data series length and recent data have an important effect on the estimation accuracy. |
| |
| Keywords: | urban water demand multicollinearity partial least squares regression model |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
