节点文献
回归误差方差的区间估计
An interval estimator of error variance in nonparametric regression model
【摘要】 首先将回归函数限制在一个有限维函数空间中,得到一个近似的线性模型,在此基础上,由Fiducial推断给出了误差方差的区间估计。该区间估计形式简单,易于计算。给出了这个区间估计的真实覆盖率和名义水平差异的一个上界,该上界由回归函数与其近似的线性函数的距离界定,并且对所给区间估计的真实覆盖率进行了数值模拟。
【Abstract】 Firstly an approximate linear model was obtained by confining the regression function in a functional space with finite dimension.Then the interval estimator was proposed by Fiducial inference.The interval estimator has a simple form and is easy to be calculated.An upper bound of the difference between the true and the nominal level of the interval estimator was given. It is dominated by the distance between the regression function and its approximating function.True percentage of coverage for the interval estimator was also simulated.
【关键词】 非参数回归;
误差方差;
区间估计;
置信水平;
【Key words】 nonparametric regression; error variance; interval estimator; confidence level;
【Key words】 nonparametric regression; error variance; interval estimator; confidence level;
【基金】 国家自然科学基金(10271013)
- 【文献出处】 中国石油大学学报(自然科学版) ,Journal of China University of Petroleum(Edition of Natural Science) , 编辑部邮箱 ,2007年04期
- 【分类号】O212.7
- 【下载频次】165