Arguably the number of different meta-heuristics \(N\) is already too large. One simple way to to make this number quickly \(N^2\) is to combine them. Here this is done with Leaping Frogs and Artificial Bees.
In order to obtain better generalization abilities and mitigate the impacts of the best and worst individuals during the process of optimization, this paper suggests Bee and Frog Co-Evolution Algorithm(abbreviation for BFCEA), which combines Mnemonic Shuffled Frog Leaping algorithm With Cooperation and Mutation(abbreviation for MSFLACM) with improved Artificial Bee Colony(abbreviation for ABC). The contrast experimental study about different iteratively updating strategies was acted in BFCEA, including strategy of integrating with ABC, regeneration of the worst frog and its leaping step. The key techniques focus on the first 10 and the last 10 frogs evolving ABC in BFCEA, namely, the synchronous renewal strategy for those winner and loser should be applied, after certain G times’ MSFLACM-running, so as to avoid trapping local optimum in later stage. The ABC evolution process will be called between all memes’ completing inner iteration and all frogs’ outer shuffling, the crossover operation is removed from MSFLACM for its little effect on time-consuming and convergence in this novel algorithm. Besides, in ABC, the scout bee is generated by Cauchy mutating instead at random. The performance of proposed approach is examined by well-known 16 numerical benchmark functions, and obtained results are compared with basic Shuffled Frog Leaping algorithm(abbreviation for SFLA), ABC and four other variants. The experimental results and related application in cloud resource scheduling show that the proposed algorithm is effective and outperforms other variants, in terms of solution quality and convergence, and the improved variants can obtain a lower degree of unbalanced load and relatively stable scheduling strategy of resources in complicated cloud computing environment.