Test Case Quality Measures in the Optimization of a Genetic Algorithm for Sorting Networks

Abstract- The sorting network problem has been of interest to the computer science community in general, and to the field of evolutionary computing, for some time. In this paper we present two measures of test case quality that may have application to other problems in which fitness is based on a large number of test cases. We use these measures to reduce the number of tests used in fitness evaluation as a means of optimizing the speed of a genetic algorithm. We report the results of several experiments which seem to lend support to the merit of these measures