罗艳虹,胡良平.适应性回归分析(Ⅱ)———排除噪声变量的干扰[J].四川精神卫生杂志,2019,32(2):101-104,109.,Adaptive regression analysis(Ⅱ)——eliminating the disturbing of the noise variables[J].SICHUAN MENTAL HEALTH,2019,32(2):101-104,109
适应性回归分析(Ⅱ)———排除噪声变量的干扰
Adaptive regression analysis(Ⅱ)——eliminating the disturbing of the noise variables
  
DOI:10.11886/j.issn.1007-3256.2019.02.002
中文关键词:  适应性  样条  回归分析  基函数  噪声变量  失拟  拟合优度  重要性
英文关键词:Adaptability  Spline  Regression analysis  Basis function  Noise variable  Lack of fit  Goodness of fit  Importance
基金项目:国家高技术研究发展计划课题资助(2015AA020102)
作者单位
罗艳虹 山西医科大学公共卫生学院卫生统计学教研室世界中医药学会联合会临床科研统计学专业委员会 
胡良平 军事科学院研究生院世界中医药学会联合会临床科研统计学专业委员会 
摘要点击次数:
全文下载次数:
中文摘要:
      【摘要】 本文目的是通过分析一个带有 8 个噪声变量的数据集,揭示适应性回归模型的实际应用价值。在数据集包含两个自变量与因变量有密切数量联系的前提条件下,适应性回归模型受噪声变量的影响接近于零;在数据集包含一个自变量与因变量有密切数量联系的前提条件下,适应性回归模型受噪声变量的影响较大,其分析结果出现了一定程度的“ 失真”;在数据集包含零个自变量与因变量有密切数量联系的前提条件下,适应性回归模型受噪声变量的影响非常大,其分析结果是完 全不可信的。得出的结论是:适应性回归分析模型不是万能的,其结果的可信度取决于数据集中是否真正包含“ 客观存在的规律性”。
英文摘要:
      This paper revealed the practical application value of the adaptive regression model through a data set with eight noise variables. Under the premise that the data set had a close quantitative relationship between two independent variables and the dependent variable, the influence of the noise variables on the adaptive regression model was close to zero. Under the preconditions that the data set had a close quantitative relationship between an independent variable and the dependent variable, the adaptive regression model was remarkably affected by the noise variables, and the results showed a certain degree of " distortion" . Under the premise that the data set did not have a close quantitative relationship between the independent variables and the dependent variable, the adaptive regression model was greatly affected by the noise variables, and the results of the analysis were completely unreliable. The conclusion was reached as follows: the adaptive regression analysis model was not universal, and the credibility of the results came from the approach mentioned above depended on whether the data set truly contained " the regularity of objective existence" .
查看全文  查看/发表评论  下载PDF阅读器
关闭