Hi,
I googled for a java package for multiobjective optimization techniques based on evolutionary algorithms (including but not limited to GA and PSO). The package should provide APIs to be used by other Java programs (e.g., Cloudsim for cloud computing simulation, Storm for multiprocessor real-time simulation). If a GUI is provided, it will be good but not necessary.
The most common results were:
ECJ (http://cs.gmu.edu/~eclab/projects/ecj/)
EVA2 (http://www.ra.cs.uni-tuebingen.de/software/EvA2/)
jMetal (http://jmetal.sourceforge.net/)
Opt4J (http://opt4j.sourceforge.net/)
The point is I'm not sure which one to use. jMetal and Opt4J mention the other frameworks in their documents, but I couldn't find papers comparing them together.
If any one has tried them, will you please advise me which one to use?
Thanks
Hi
One important aspect, must which be considered for long-term utilization of the open-source packages is the rate of the updating the versions of the project to keep it live in view of the being progressed for most up-to-date changes of the related science and their implementation within the aforementioned distributions...
With due attention to above case, one can assert that jMetal's path to the maintenance and permanent updating is more honorable than the others... whereas the functionality doesn't annihilate the others...
Best
Hi Mohammed
I'm studying the multiobjective optimization algorithms in CloudSim, as your question mentioned before. the jMetal can be used in CloudSim. May you show me some examples using it in CloudSim? Tanks in advance.
Hi Maple
I haven't yet written any code for CloudSim. I just wanted to decide which tool to use for writing code. Currently, I'm reading in Opt4J and jMetal. I think I'll choose the one that provides:
- Badges
- Science method