Abstract: In order to solve the problem that ant colony optimization is easily trapped into local optimum, a new algorithm called enhanced ant colony optimization based on genetic algorithm was proposed. It maintained a history best optimal solution which was combined with part of current best solutions as the initial chromosomes during iterations, and adopted genetic algorithm to search better solutions at a larger space. It selected the current best solution from results of ants and chromosomes to prevent ant colony optimization from being trapped into local optimum. It was tested on feature selection problem of binary classification and compared with particle swarm optimization and differential evolution, the experiments were implemented on three benchmark datasets, and the results show the efficiency and superiority of the proposed algorithm.