Solving the Attribute Reduction Problem with Ant Colony Optimization

Attribute reduction is an important process in rough set theory. More minimal attribute reductions are expected to help clients make decisions in some cases, though the minimal attribute reduction problem (MARP) is proved to be an NP-hard problem. In this paper, we propose a new heuristic approach for solving the MARP based on the ant colony optimization (ACO) metaheuristic. We first model the MARP as finding an assignment which minimizes the cost in a graph. Afterward, we introduce a preprocessing step that removes the redundant data in a discernibility matrix through the absorption operator and the cutting operator, the goal of which is to favor a smaller exploration of the search space at a lower cost. We then develop a new algorithm R-ACO for solving the MARP. Finally, the simulation results show that our approach can find more minimal attribute reductions more efficiently in most cases.