Going the Extra Mile in Ant Colony Optimization Experimental Evaluation

Ant colony optimization (ACO) is a novel and promising meta-heuristic for solving hard combinatorial optimization problems, such as travelling salesman[3] [4] [5] [6] [7] [8], quadratic assignment[9] [10] [11] [12] and set covering problem [13]. Unfortunately, according to systems performance evaluation literature, the methods adopted to experimentally validate the current ACO are not enough accurate and are excessively time-consuming. In fact, some works on heuristic optimization, suffer from the same methodological limitations. However, since the research on ant colony optimization is recent, this community has the opportunity of rethinking the evaluation methods from scratch, avoiding misconception on the evaluation the experimental data.