Novel fruit fly optimization algorithm with trend search and co-evolution

To solve both continuous function optimization and clustering parameter problems, the novel fruit fly optimization algorithm with trend search and co-evolution (CEFOA) was proposed. It is featured with several mechanisms devised for solving the concerned problems: 1) trend search strategy was proposed and embedded into FOA. The strategy consisted two steps, which were multidimensional food evaluation method and trend search. Multidimensional food evaluation method was introduced to estimate the quality of the food sources. In the basic of the proposed method, trend search was applied to enhance the local searching capability of fruit fly swarm; 2) co-evolution mechanism was employed to avoid the premature convergence and improve the ability of global searching. To verify the performance of CEFOA we tested 26 benchmark functions with different characteristic. Experimental results indicated that CEFOA had better precision and convergence speed than several other swarm intelligence algorithms. In addition, it is applied to enhance the clustering precision and efficiency. We utilized the improved model to optimize the parameter p in Affinity Propagation clustering (AP). The simulation results demonstrated that AP clustering algorithm with CEFOA was prior to AP clustering algorithms with CLPSO, BLPSO and IFOA, which were the top three algorithms in precious tests. The new clustering model had more robust without setting parameter manually. Thus, the proposed algorithm had a better research potential and a good application value.