作者: B.F. Jones , H.-H. Sthamer , D.E. Eyres
关键词: Test data generation 、 Integration testing 、 Model-based testing 、 Non-regression testing 、 Algorithm 、 White-box testing 、 Keyword-driven testing 、 Fitness function 、 Computer science 、 Manual testing
摘要: Genetic algorithms have been used to generate test sets automatically by searching the domain of software for suitable values satisfy a predefined testing criterion. These criteria set requirements data adequacy structural testing, such as obtaining full branch coverage and controlling number iterations conditional loop. This technique has applied successfully several problems, varying in complexity from quadratic equation solver generic sort module that comprises procedures. In these cases, was obtained. could be approaches other than provided goal is clearly defined, fitness function which relates this can devised give single numeric value fitness. The quality enhanced designing close subdomain boundary where likelihood revealing an error higher.