Abstract
In this work we evaluate the usefulness of a Path Relinking based method for generating the initial population of Multi-Objective Genetic Algorithms and evaluate its performance on the Multi-Objective Next Release Problem.The performance of the method was evaluated for the algorithms MoCell and NSGA-II, and the experimental results have shown that it is consistently superior to the random initialization method and the extreme solutions method, considering the convergence speed and the quality of the Pareto front, that was measured using the Spread and Hypervolume indexes.
Author supplied keywords
Cite
CITATION STYLE
Da Silva, T. G. N., Rocha, L. S., & Maia, J. E. B. (2014). An effective method for MOGAs initialization to solve the multi-objective next release problem. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8857, pp. 25–37). Springer Verlag. https://doi.org/10.1007/978-3-319-13650-9_3
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
