Test Case Prioritization with Genetic Algorithms for Regression Testing

Abstract

Regression testing can become time-consuming when large test suites are executed after every software modification without considering fault-detection order. This article presents a Genetic Algorithm-based test case prioritization framework for improving regression testing efficiency. The framework represents each candidate test order as a permutation chromosome and applies fitnessbased selection, order-preserving crossover, swap or insertion mutation, and termination control to evolve stronger execution sequences. The fitness function combines historical fault detection, code coverage, requirement criticality, recent change relevance, and execution time so that high-value tests are placed earlier without overloading the beginning of the suite with longrunning cases. The study shows that Genetic Algorithm prioritization improves early fault discovery, reduces the time required to detect critical failures, and provides a more balanced ordering than original, random, historical-failure, or coverageonly strategies. The proposed approach supports faster regression feedback, better use of testing resources, and more efficient maintenance of changing software systems.