Improving the Boundaries of Sequential Statistic Criteria by Computer Simulation

Abstract

The effectiveness of sequential criteria in tests of two simple hypotheses is investigated. An algorithm for calculating the improved boundaries of Wald-type sequential criteria based on the likelihood ratio is proposed. The proposed algorithm makes it possible to find the boundaries with given precision and lesser computational costs by comparison with previously known algorithms. A software program by means of which improved estimates of the boundaries of the criteria may be calculated in the least number of iterations is developed. The effectiveness of the proposed algorithm is demonstrated in concrete examples.