A Virus Evolutionary Genetic Algorithm Using Local Selection

Virus evolutionary genetic algorithm (VEGA) is an improved genetic algorithm (GA) that can prevent premature convergence, which introduces an additional virus population and two infection operators to GA. In this paper, a VEGA using the local selection scheme is proposed. Local selection can effectively maintain the diversity of the host population in VEGA and then improved the algorithm´s performance. Survivals of local selection are with high fitness, and this indirectly leads to the elimination of the virus individuals that contain ineffective schemata, which partly suppresses the large iteration time due to the transduction operator of VEGA. The experimental result shows the effectiveness of the proposed algorithm.