On the Quasi-Stationary Distribution of the Shiryaev-Roberts Diffusion
arXiv:1606.06658 · doi:10.1080/07474946.2016.1275512
Abstract
We consider the diffusion generated by the equation with fixed, and where is given, and is standard Brownian motion. We assume that is stopped at with preset, and obtain a closed-from formula for the quasi-stationary distribution of , i.e., the limit , . Further, we also prove to be unimodal for any , and obtain its entire moment series. More importantly, the pair with and is the well-known Generalized Shiryaev-Roberts change-point detection procedure, and its characteristics for are of particular interest, especially when is large. In view of this circumstance we offer an order-three large- asymptotic approximation of valid for all . The approximation is rather accurate even if is lower than what would be considered "large" in practice.
arXiv comment25 pages; 2 figures
References in corpus (4)
- State-of-the-Art in Sequential Change-Point Detection
- On optimality of the Shiryaev-Roberts procedure for detecting a change in distribution
- An Analytic Expression for the Distribution of the Generalized Shiryaev-Roberts Diffusion
- Exact Distribution of the Generalized Shiryaev-Roberts Stopping Time Under the Minimax Brownian Motion Setup