Distributed optimization on super computers
Ramin Etemaadi, Michel R. V. Chaudron · 2014
Nowadays advanced software systems need to satisfy large number of quality attributes at the same time. It is a very complex optimization problem which software architects must address. Evolutionary algorithms can help architects to find optimal solutions which meet these conflicting quality attributes. However, these metaheuristic approaches in multiobjective problems especially for high dimensions mostly take so long time to be executed. One of the best solutions to speed up this process is distributing execution of evolutionary algorithm on multiple nodes of a super computer or on the cloud.