A Bias Bound Approach to Non-parametric Inference

A Bias Bound Approach to Non-parametric Inference
复制标题

非参数推理的偏差约束方法

DOI:
10.1093/restud/rdz065
复制
发表时间:
2020
期刊:
The Review of Economic Studies
影响因子:
--
通讯作者:
Schennach, Susanne M
Schennach, Susanne M
中科院分区:
--
文献类型:
--
作者:
Schennach, Susanne M

文献摘要

参考文献

被引文献

相似文献

传统的方法是选择一个平滑参数,使得估计量的偏差相对于其标准差是可以忽略的,以获得非参数的有效可信区间。虽然这种方法看起来很简单,但它有两个缺点:第一,最佳带宽选择的问题不再是明确的,因为不清楚偏差与标准偏差的比率应该被认为是可以忽略的。其次,由于带宽选择必然偏离最优(均方最小化)带宽,这样的置信度区间是非常低效的。为了解决这些问题,我们构造了有效的可信区间来解释不可忽略的偏差的存在,从而使得在最优均方误差最小化带宽的情况下进行推断成为可能。实现这一点的关键困难是找到一个严格但可行的关于非参数估计偏差的界限。众所周知,不可能一致地估计最优非参数估计量的逐点偏差(否则,人们可以减去它而获得更快的收敛速度,违反了Stone关于最优收敛速度的界)。然而,我们发现,在最小的原始假设下,一致地估计偏差大小的上界是可能的,这足以提供一个有效的可信区间,其长度以最优速度递减,并且与Stone的结果不矛盾。
The traditional approach to obtain valid confidence intervals for non-parametric quantities is to select a smoothing parameter such that the bias of the estimator is negligible relative to its standard deviation. While this approach is apparently simple, it has two drawbacks: first, the question of optimal bandwidth selection is no longer well-defined, as it is not clear what ratio of bias to standard deviation should be considered negligible. Second, since the bandwidth choice necessarily deviates from the optimal (mean squares-minimizing) bandwidth, such a confidence interval is very inefficient. To address these issues, we construct valid confidence intervals that account for the presence of a non-negligible bias and thus make it possible to perform inference with optimal mean squared error minimizing bandwidths. The key difficulty in achieving this involves finding a strict, yet feasible, bound on the bias of a non-parametric estimator. It is well-known that it is not possible to consistently estimate the pointwise bias of an optimal non-parametric estimator (for otherwise, one could subtract it and obtain a faster convergence rate violating Stone’s bounds on the optimal convergence rates). Nevertheless, we find that, under minimal primitive assumptions, it is possible to consistently estimate anupper boundon the magnitude of the bias, which is sufficient to deliver a valid confidence interval whose length decreases at the optimal rate and which does not contradict Stone’s results.
密度导数估计中的平滑偏差
DOI: --
发表时间: 1993
期刊:
影响因子: --
作者:
Thomas M. Stoker
通讯作者: Thomas M. Stoker
使用弱可分离性的非参数回归估计
DOI: --
发表时间: 2001
期刊:
影响因子: --
作者:
J. Pinkse;John W. Galbraith;D. Green;N. Heckman;J. Horowitz;Oliver Linton;Rosa L. Matzkin;P. Robinson;M. Slade
通讯作者: M. Slade
DOI: --
发表时间: 2013
期刊:
影响因子: --
作者:
T. Cai;Mark G. Low;Yin Xia
通讯作者: Yin Xia
DOI: 10.3982/ecta14434
发表时间: 2018-03-01
期刊: ECONOMETRICA
影响因子: 6.1
作者:
Armstrong, Timothy B.;Kolesar, Michal
通讯作者: Kolesar, Michal
DOI: 10.1080/01621459.2015.1017578
发表时间: 2015-12-01
影响因子: 3.7
作者:
Calonico, Sebastian;Cattaneo, Matias D.;Titiunik, Rocio
通讯作者: Titiunik, Rocio