A new method to solve the “Boundary Effect” of Adaptive random testing

摘要

Compare to random testing, Adaptive random testing generates test cases evenly in the input domain so that it can detect the first failure of the program using fewer test cases. But It is found that adaptive random testing will be more inclined to generate test cases on the boundary in the input domain, which is called “Boundary Effect”. The “Boundary Effect” will affect the performance of the adaptive random testing in certain situations, such as in multi-dimensional input domain. This paper presents an effective way to solve the “Boundary Effect” of adaptive random testing in order to enhance the performance in high-dimensional input domain.