Estimating a multivariate normal mean with a bounded signal to noise ratio under scaled squared error loss

Othmane Kortbi, Éric Marchand

Research output: Contribution to journalArticlepeer-review


For normal models with X ~ Np(θ, σ2 Ip), S22χ2 k, independent, we consider the problem of estimating θ under scale invariant squared error loss ||d − θ||22, when it is known that the signal-to-noise ratio ||θ||/σ is bounded above by m. Risk analysis is achieved by making use of a conditional risk decomposition and we obtain in particular sufficient conditions for an estimator to dominate either the unbiased estimator δUB(X) = X, or the maximum likelihood estimator δML(X,S2), or both of these benchmark procedures. The given developments bring into play the pivotal role of the boundary Bayes estimator δBU,0 associated with a prior on (θ,σ2) such that θ|σ2 is uniformly distributed on the (boundary) sphere of radius mσ and a non-informative 1/σ2 prior measure is placed marginally on σ2. With a series of technical results related to δBU,0; which relate to particular ratios of confluent hypergeometric functions; we show that, whenever m ≤ √p and p ≥ 2, δBU,0 dominates both δUB and δML. The finding can be viewed as both a multivariate extension of p = 1 result due to Kubokawa (Sankhyā 67:499–525, 2005) and an unknown variance extension of a similar dominance finding due to Marchand and Perron (Ann Stat 29:1078–1093, 2001). Various other dominance results are obtained, illustrations are provided and commented upon. In particular, for m ≤ √p/2, a wide class of Bayes estimators, which include priors where θ|σ2 is uniformly distributed on the ball of radius mσ centered at the origin, are shown to dominate δUB.

Original languageEnglish
Pages (from-to)277-299
Number of pages23
JournalSankhya A
Issue number2
Publication statusPublished - Aug 1 2013
Externally publishedYes

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty


Dive into the research topics of 'Estimating a multivariate normal mean with a bounded signal to noise ratio under scaled squared error loss'. Together they form a unique fingerprint.

Cite this