Modified ACO to maintain diversity in regression test optimization

Regression testing is unavoidable maintenance activity that is performed several times in software development life cycle. Optimization of regression test case is required to minimize the test case (which will in-turn reduce the time and cost of testing) and to find the fault in early testing activity. The two widely used regression test case optimization techniques, namely, selection and prioritization are recently found to be integrated with different metaheuristic algorithms for fruitful regression test cases. Among the various meta-heuristic algorithms, Ant colony optimization (ACO) algorithm is most popularly used. ACO will try to find the smallest path out all the test cases and it is not sufficient because it will not cover all the test cases which are needed. In this paper we have proposed a modified ant colony optimization to solve test cases in huge search space. The modified algorithm selects the best test cases that find the maximum fault in minimum time.