Comparative Analysis of Test Case Prioritization Using Ant Colony Optimization Algorithm and Genetic Algorithm

Abstract

After it is published, every software system will get an upgrade, requiring it to adapt to meet the ever-changing client needs thus regression testing becomes one of the most important operations in any software system. As it is too expensive to repeat the execution of all the test cases available from a previous version of the software system, numerous ways to optimize the regression test suite have evolved, one of which is test case prioritizing. This study was carried out to test and compare the effectiveness of evolutionary algorithms and swarm intelligence algorithms, represented by the Genetic Algorithm (GA) and Ant Colony Optimization (ACO) algorithms. They will be implemented to find the Average Percentage Fault Detected (APFD), execution time, and Big O notation, as these are critical aspects in software testing to ensure high-quality products are produced on time. This study employs data from two separate investigations on comparable issues, denoted as Case Study One and Case Study Two. TCP has been extensively used in recent years, but not much research has been conducted to analyze and evaluate the performance of genetic algorithms (GA) and ant colony optimization (ACO) in a test case prioritization context. The algorithms were compared using APFD, execution time. The APFD and execution time values of 50, 100, 150, and 200 iterations for ACO and GA for both datasets are conducted. Both algorithms were determined to work on O() notation, which indicates they should scale up their execution process similarly on different input scales. Both algorithms performed well in their respective roles. ACO has shown to be more valuable than GA in terms of APFD and GA has shown to be more valuable than ACO in terms of execution time.