Dynamic sequential team multi-hypothesis testing under uniformly distributed nonstationary observations

We consider a dynamic team of communicating decision makers performing Bayesian sequential multi-hypothesis testing under nonstationary observations. A simple decision rule is shown to be member-by-member optimal if the observations are uniformly distributed, if the observation costs are defined to reflect the quality of observations, and if “vague” terminal decisions are allowed without additional penalty. The “vague” terminal decisions can be made “clear” by performing a finite number of member-by-member optimal sequential decisions successively.