Research on and application of FOA based on dynamic population and direction correcting

Fruit fly Optimization Algorithm (FOA) is a kind of relatively new swarm intelligence optimization algorithm with strong performance. In this paper a kind of FOA based on dynamic population and direction correct (DPDC-FOA) is proposed on the question of premature convergence. By changing the fruit fly group's search range, number of individuals and group position selection strategy during the movement process, this improved algorithm enhances the basic algorithm's ability to search the global optimal. We use 6 functions as benchmark and the result shows that the improved algorithm has better performance in the rate and the precision of convergence. Using this algorithm in the parameter optimization of support vector machine (SVM), the classification accuracy got increased.