Mirror adaptive random testing

Adaptive random testing (ART) has recently been introduced to improve the fault-detection effectiveness of random testing (RT) for certain types of failure-causing patterns. However, ART requires extra computations to ensure an even spread of test cases, which may render ART to be less cost-effective than RT. In this paper, we introduce an innovative approach, namely Mirror Adaptive Random Testing (MART), to reduce these computations. Our simulation results clearly show that MART does improve the cost-effectiveness of ART.