A Markov Decision Approach to Optimize Testing Profile in Software Testing

In this paper, we demonstrate an approach to optimize software testing, minimize the expected cost with given software parameters of concern. Taking software testing process as a Markov decision process, a Markov decision model of software testing is proposed, and by using a learning strategy based on the Cross-Entropy method to optimize the software testing, we obtain the optimal testing profile. Simulation results show that this learning strategy reduces significantly in expected cost comparing with random testing, moreover, this learning strategy is more feasible and significantly in reducing the number of test cases required to detect and revealing a certain number of software defects than random testing.