Let be a discrete-time stochastic process with a distribution , , where is an open subset of the real line. We consider the problem of testing a simple hypothesis versus a composite alternative , where is some fixed point. The main goal of this article is to characterize the structure of locally most powerful sequential tests in this problem. For any sequential test with a (randomized) stopping rule and a (randomized) decision rule let be the type I error probability, the derivative, at , of the power function, and an average sample number of the test . Then we are concerned with the problem of maximizing in the class of all sequential tests such that \alpha(\psi,\phi)\leq \alpha\quad{and}\quad \mathscr N(\psi)\leq \mathscr N, where and are some restrictions. It is supposed that is calculated under some fixed (not necessarily coinciding with one of ) distribution of the process . The structure of optimal sequential tests is characterized.
View on arXiv