TY - JOUR A2 - Van Hulle, Marc AU - Villagra, Andrea AU - Alba, Enrique AU - Leguizamón, Guillermo PY - 2016 DA - 2016/06/14 TI - A Methodology for the Hybridization Based in Active Components: The Case of cGA and Scatter Search SP - 8289237 VL - 2016 AB - This work presents the results of a new methodology for hybridizing metaheuristics. By first locating the active components (parts) of one algorithm and then inserting them into second one, we can build efficient and accurate optimization, search, and learning algorithms. This gives a concrete way of constructing new techniques that contrasts the spread ad hoc way of hybridizing. In this paper, the enhanced algorithm is a Cellular Genetic Algorithm (cGA) which has been successfully used in the past to find solutions to such hard optimization problems. In order to extend and corroborate the use of active components as an emerging hybridization methodology, we propose here the use of active components taken from Scatter Search (SS) to improve cGA. The results obtained over a varied set of benchmarks are highly satisfactory in efficacy and efficiency when compared with a standard cGA. Moreover, the proposed hybrid approach (i.e., cGA+SS) has shown encouraging results with regard to earlier applications of our methodology. SN - 1687-5265 UR - https://doi.org/10.1155/2016/8289237 DO - 10.1155/2016/8289237 JF - Computational Intelligence and Neuroscience PB - Hindawi Publishing Corporation KW - ER -