An Analytical Shrinkage Estimator for Linear Regression

Lassance, Nathan
(2023) Statistics & Probability Letters — Vol. 194, p. 109760 (2023)

Files

LFIN_RP_2023-03.pdf
  • Open Access
  • Adobe PDF
  • 523.78 KB
Download

Details

Authors
Abstract
We derive an analytical solution to the optimal shrinkage of OLS regression coefficients toward a constant target, under any first two moments of predictors. The estimator closely mimics the prediction performance of ridge penalty, which admits no general analytical solution.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Lassance, N. (2023). An Analytical Shrinkage Estimator for Linear Regression. Statistics & Probability Letters, 194, 109760. https://doi.org/10.1016/j.spl.2022.109760 (Original work published 2023)