A general theory of purely sequential minimum risk point estimation (MRPE) of a function of the mean in a normal distribution. Issue 4 (2nd October 2019)

• Record Type: Journal Article
• Authors: Mukhopadhyay, Nitis; Wang, Zhe
• Is Part Of: Sequential analysis. Volume 38:Issue 4(2019)
• Page Start: 480
• Page End: 502
• Publication Date: 2019-10-02
• Subjects: Asymptotic efficiency -- asymptotic risk efficiency -- bias correction -- first-order properties -- illustrations -- linear cost -- minimum risk -- optimal fixed sample size -- regret -- regret expansion -- risk efficiency -- second-order properties -- squared error loss (SEL) -- uniform integrability -- weighted SEL; 62L12 -- 62L10 -- 62L05; Sequential analysis -- Periodicals; 519.54
• DOI: 10.1080/07474946.2019.1686885
• Languages: English
• ISSNs: 0747-4946

Abstract

Abstract: A purely sequential minimum risk point estimation (MRPE) methodology with associated stopping time N is designed to come up with a useful estimation strategy. We work under an appropriately formulated weighted squared error loss (SEL) due to estimation of g ( μ ), a function of μ, with g ( X ¯ N ) plus linear cost of sampling from a N ( μ, σ 2 ) population having both parameters unknown. A series of important first-order and second-order asymptotic (as c, the cost per unit sample, → 0 ) results is laid out including the first-order and second-order efficiency properties. Then, accurate sequential risk calculations are launched, which are then followed by two main results: (i) Theorem 4.1 shows an asymptotic risk efficiency property, and (ii) Theorem 5.1 shows an asymptotic second-order regret expansion associated with the proposed purely sequential MRPE strategy assuming suitable conditions on g (.). We also provide a bias-corrected version of the terminal estimator, g ( X ¯ N ) . We follow up with a number of interesting illustrations where Theorems 4.1–5.1 are readily exploited to conclude an asymptotic risk efficiency property and second-order regret expansion, respectively. A number of other interesting illustrations are highlighted where it is possible to verify the conclusions from Theorems 4.1–5.1 more directly with less stringent assumptions on the pilot sample size.