Stein's lemma relates Gaussian multiplication to differentiation. Let . One sufficient set of conditions is that is locally absolutely continuous, both and are finite, and as , where is the standard normal density. ThenSince , integration by parts on a finite interval moves the derivative from to ; the boundary condition and integrability justify the infinite-interval limit.
For , the identity gives . For , it gives . More generally, for with , standardization yieldsunder the corresponding conditions. The factor is essential.
This integration-by-parts rule converts Gaussian error cross terms into derivatives of an estimator, which is the mechanism of the Stein risk identity for normal mean estimators. Boundary behavior matters for singular estimators: formal differentiation alone does not justify applying the identity across a nonintegrable singularity.
For , expand the squared Estimation error and apply Gaussian integration by parts to each cross term. When satisfies the integrability and boundary conditions needed for that identity, this gives the displayed formula. For and , the origin singularity is integrable and the inner boundary term is , so the same identity applies by excision. Its divergence is , giving risk .
Articles by others on the same topic
There are currently no matching articles.