摘要
本文利用多元线性回归方法对北京市自来水总销售量数据进行建模。用主成分分析产生以北京市国民生产总值、常住人口数、污水处理率、年降水量、城市绿化覆盖率为原始变量的线性组合作为主成分,而后利用比较标准,建立"最优"的回归方程。模型通过残差图检验可知回归效果良好,具有预测能力。本文的建模方法以及模型为较准确地预测北京市自来水总销售量提供参考。
In this paper, the use of multiple linear regression method of Beijing Waterworks total sales data modeling. Beijing produced GDP, the resident population, the sewage treatment rate, annual rainfall, urban green coverage to a linear set of original variables of cooperation as the main ingredient, then the use of comparative standards, establish"optimal"principal component analysis of the regression equation. Model tested by residual plot shows that the return to good effect, with the predictive ability. Modeling methods and model reference herein to more accurately predict the Beijing Waterworks total sales.
出处
《科教导刊》
2015年第12期167-168,190,共3页
The Guide Of Science & Education
关键词
自来水总销售量
主成分分析
多元线性回归
R语言
total sales of tap water
principal component analysis
multiple linear regression
R language
