Ant colony optimization for feature selection via a filter-randomized search heuristic.

The combination of bio-inspired algorithms with the design of new search heuristics has allowed the development of innovative methods for combinatorial problems where the number of possible solutions is unaffordable. However, for years swarm intelligent methods have proven to be adequate to deal with these issues. Therefore, this paper proposes a new application of the Ant Colony Optimization (ACO) algorithm for Feature Selection (FS). Specifically, the proposed algorithm applies ACO for FS using a filter approach as a heuristic and makes use of a randomized search, resulting in a very innovative algorithm for variable reduction.