Answers are grounded in this paper and cite their sources.
Suppose X₁, X₂, … are independent random variables such that for some unknown ν, each of X₁, …, Xν - 1 is distributed according to F₀, while X_ν, Xν + 1, … are all distributed according to F₁. We prove a result of Moustakides that claims that the CUSUM procedures are optimal in the sense of Lorden. We do that by proving that the procedures are Bayes for some stochastic mechanism of generating ν.
No takes yet. Share an insight, caveat, or question.
Ya’acov Ritov (1990) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: