Ant Colony Optimization Algorithm Parameter Tuning for T-way IOR Testing

In this study, Ant Colony Optimization (ACO) algorithm's parameters for t-way IOR testing were examined. ACO and its variant have been applied to t-way testing but never to t-way IOR interaction support. Tuning ACO parameters were executed to ensure that ACO could perform for IOR as good as other t-way interaction support. Parameter alpha, beta, tau(0), q(0), rho value and number of ant were tuned to uniform and non-uniform configuration. Each parameter was executed for 10 independent run. Average best test suite and best test suite were recorded and compared among other parameter values to find which value will produce the best result. The optimum test suite size and average test suite size were generated when the value for parameter alpha = 0.5, beta = 3, tau(0) = 0.4 (uniform configuration) and, 0.2 and 1 (non-uniform configuration), q(0) = 0.5, rho = 0.5 and number of ant = 20.