Stability of the posterior mean in linear models An admissibility property of D-optimum and E-optimum designs

✍ Scribed by Marek Męczarski; Ryszard Zieliński

• Book ID: 104302499
• Publisher: Elsevier Science
• Year: 1997
• Tongue: English
• Weight: 301 KB
• Volume: 33
• Category: Article
• ISSN: 0167-7152
• DOI: 10.1016/s0167-7152(96)00118-6

✦ Synopsis

Stability (robustness) of the Bayes estimates in linear models is considered. As measures of stability, the volume and the diameter of the ellipsoid of the posterior mean are proposed. The classes of priors of interest are sets of normal probability distributions with varying covariance matrices. The admissibility property of D-optimum and E-optimum designs under an optimality concept generated by the robustness idea is stated.