Estimation of linear projections of non-sparse coefficients in high-dimensional regression

Estimation of linear projections of non-sparse coefficients in high-dimensional regression
复制标题

高维回归中非稀疏系数线性投影的估计

DOI:
--
复制
发表时间:
2020
影响因子:
1.1
通讯作者:
A. Schwartzman
A. Schwartzman
中科院分区:
数学3区
文献类型:
--
作者:
David Azriel;A. Schwartzman

文献摘要

参考文献

被引文献

相似文献

:在这项工作中,我们研究的参数的数量远远大于观察的数量时,信号的估计。大量的文献假设这类问题的稀疏结构,其中大多数参数为零或接近零。当这个假设不成立时,我们可以关注参数向量的低维函数。在这项工作中,我们研究一维线性投影。具体来说,在高维线性回归的背景下,感兴趣的参数是β,我们研究T β的估计。我们证明了当p > n时,T β是极小极大且可容许的,其中T β是最小二乘估计,使用伪逆.因此,对于线性投影,不需要正则化或收缩。该估计量易于分析,并且可以构造置信区间。我们研究了来自脑成像的高维数据集,其中显示信号是弱的,非稀疏的,并且与零显著不同。
: In this work we study estimation of signals when the number of parameters is much larger than the number of observations. A large body of literature assumes for these kind of problems a sparse structure where most of the parameters are zero or close to zero. When this assumption does not hold, one can focus on low-dimensional functions of the parameter vector. In this work we study one-dimensional linear projections. Specifi-cally, in the context of high-dimensional linear regression, the parameter of interest is β and we study estimation of a T β . We show that a T ˆ β , where ˆ β is the least squares estimator, using pseudo-inverse when p > n , is minimax and admissible. Thus, for linear projections no regularization or shrinkage is needed. This estimator is easy to analyze and confidence intervals can be constructed. We study a high-dimensional dataset from brain imaging where it is shown that the signal is weak, non-sparse and significantly dif- ferent from zero.
DOI: 10.1093/biomet/asq075
发表时间: 2011-03-01
期刊: BIOMETRIKA
影响因子: 2.7
作者:
Schwartzman, Armin;Lin, Xihong
通讯作者: Lin, Xihong