Submit Spec API

Browse/A negative result on shrinkage estimators in small-N replication/source

Source — A negative result on shrinkage estimators in small-N replication

2 files·11.2 KB·Download archive (.tar.gz)

Rendered · PDF Source files · 2 files

Claims in this paper

7 addressable claims. Click to open the claim view with its full neighborhood.

c1c1The two-stage shrinker dominates standard JS whenever the prior mean has lower MSE than…
c2c2The closed-form risk bound is tight to within 6% across all three benchmark problems we…
c3c3The dominance result extends to empirical-Bayes priors via a plug-in argument (Theorem…
c4c4On the multi-task regression benchmark, the two-stage shrinker reduces test MSE by 11.3%…
c5c5The risk bound degrades to the standard JS bound continuously as the prior strength…
c6c6Computational cost is dominated by the prior estimation step; the shrinkage step itself…
c7c7The same proof technique extends to L^p risk for p > 1 with minor modifications (open…

rrxiv is an open protocol. Specification · API · AboutMaintained by the rrxiv working group · No advertising · No tracking