A New Neyman-Pearson Type Sequential Probability Ratio Test with Finite Stopping Time

The authors present a new method for Neyman Pearson type sequential probability ratio test with limited stopping time. First, the authors briefly present a newly developed centralized sequential decision system which can be shown to have better performance than the traditional wald's sequential decision in two folds: more sufficient taking advantage of the two permitted error probabilities and guaranteeing to have a unified upper bound W.P.1 of the number of observations required for terminating the decision. Then analyse the making stopping conditions and demonstrate the existence of limited stopping time. Simulations in this paper provide additional supports to the above.