Nasser R. Sabar‘s (Queensland University of Technology, Brisbane) and Aldeida Aleti‘s (Faculty of Information Technology, Monash University, Melbourne) paper “An Adaptive Memetic Algorithm for the Architecture Optimisation Problem” [1] describes the combination of a local search and a genetic algorithm to generate memetic algorithms.
They implement an adaptive scheme to address problems caused by the trade-off between exploration and exploitation. The tuning of the hyper-parameter and parameter values was performed using a sequential parameter optimisation technique [2]. A problem from the design of embedded systems (“component deployment”) is used to evaluate the performance. The test function tool is available online [3]. Results indicate that the proposed memetic algorithm (AMA) performed better than the single algorithms.
References
[1] N. R. Sabar and A. Aleti. An adaptive memetic algorithm for the architecture optimisation problem. In M. Wagner, X. Li, and T. Hendtlass, editors, Artificial Life and Computational Intelligence: Third Australasian Conference, ACALCI 2017, Geelong, VIC, Australia, January 31 – February 2, 2017, Proceedings, pages 254–265, Cham, 2017. Springer International Publishing.
[2] Bartz-Beielstein, T., Lasarczyk, C., Preuss, M.: Sequential parameter optimization.
In: IEEE Congress on Evolutionary Computation, pp. 773–780. IEEE (2005). DOI: 10.1109/CEC.2005.1554761