Performance and robustness in sequential testing of hypotheses

Abstract

The problem of sequential testing of parametric hypotheses is considered. Different models of data are analyzed for simple and also for complex hypotheses setting. The approaches to performance characteristics (error probabilities and expected sample sizes) calculation and to robustness analysis (under deviations from the hypothetical model assumptions) of the sequential tests are developed. Within these approaches, asymptotic expansions (w.r.t. the discretization parameter and the distortion levels) of the performance characteristics are obtained, and the robustified sequetnial procedures are constructed