statistics - Bootstrapping on Post-Lasso standard errors -
for research purposes i'm trying predict transfer prices of football players. i've implemented several models of 1 lasso. now, besides prediction, measure actual effects of variables, i've implemented post-lasso estimation (least squares on non-zero coefficients chosen lasso). i've read on several sites, , 1 here: https://www.ncbi.nlm.nih.gov/pmc/articles/pmc4207812/ standard errors wrong, because not take account selection procedure i've used before least squares estimation.
now wondering whether or not bootstrap method: bagging [1] works deal problem. sample replacements multiple times, run least squares on these samples , calculate standard errors on coefficients.
my question: work , why?
thanks lot!
[1] breiman, l. (1996). bagging predictors. machine learning, 24(2), 123-140.
Comments
Post a Comment