Optimal sequential sampling plans
Part of the Lecture Notes in Statistics book series (LNS, volume 79)
Statistical decision procedures where the sizes of sub-samples are determined sequentially (sequentially planned statistical decision procedures) are described in definition (1.22) by a sequential plan r and a terminal decision procedure ε. I.e. similarly to purely sequential decision procedures (see definition (1.12)) such a procedure is split up into
- a part τ which is focused upon the sequential determinations of sub-samples and
- a part ϕ = (ϕ a )a∈A which exclusively concerns the terminal decisions.
KeywordsDecision Procedure Sampling Plan Optimal Sampling Transition Kernel Finite Horizon
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
Unable to display preview. Download preview PDF.
- 22.There is no relation to the admissibility concept of decision theory; existence only of the expected value in the wide sense is assumed, i.e. EZ+ < ∞ or EZ- < ∞.Google Scholar
- 23.This assumption is e.g. fulfilled if Za ≤ 0 (i.e. allow an interpretation as losses); comp. Chapter 4.Google Scholar
- 24.We always define supθ … = maxθ… =-∞.Google Scholar
© Springer-Verlag New York, Inc. 1993