Probably very little, as it is a metaheuristic, with the evident characteristics of (a) not being able to take into account any profitable problems properties - such as differentiability and/or convexity - and in addition no natural stopping criteria that have anything to do with the problem statement and any characterizations of optimality. But I suppose that that is a general statement for basically all such methods, so it is probably as good, or as bad, as the rest of them.
It's a boring as watching paint dry.
I think you need to define first what you mean by other existing optimization techniques.
I believe that if scientist are today using memetic optimization techniques is because classical gradient or convex optimization techniques have failed or are limited in their performance for certain optimization problems.
I should think that the reason is that they have not adopted them well enough to the problem at hand, or even misunderstood them. The classical methods you mention are limited to continuous problems, while there is a wide spectrum of integer/combinatorial problems and methods.
Dear Venkat Rao Pasupuleti ,
I recommend to read this interesting position paper first:
Article Metaheuristics -- the metaphor exposed
To pursue research on metaheuristics and face the critics, it is essential nowadays to unify existing methods and analyses under simple technical names (e.g., local search, neighborhood, hill climbing, parameter adaptation, perturbation, restart, decision-space decomposition or projection) and focus on important complexity results, instead of wasting critical time (and reputation) by creating pointless metaphores and research societies dedicated to cats, birds, frogs and predating water drops. This being said, some of the earlier answers listed in this topic seem excessive.
Thanks Thibaut,
I just saw this after answering two other threads.
"As we have too many threads in this spirit, let me quote from one of those others:
A friend just told me not to answer (even if it is only now and then). He supposed that even good answers do not change people's mind. Anyway, we had these discussions again and again. But these discussions seem to get lost every now and then.
...
Like in every field there may be good research and bad research. So, do not put all of it together into one pot with possibly even wrong comprehensions.
Regarding the metaheuristics community, there is some awareness at least in parts of it that not everything that seems shiny is shining. A good reference to explore this view" was just given by Thibaut:
Metaheuristics—the metaphor exposed
https://doi.org/10.1111/itor.12001
"And another good pointer is to the word matheuristics which we framed to investigate the interoperation of metaheuristics and mathematical programming techniques."
https://en.wikipedia.org/wiki/Matheuristics
Have fun exploring and learning how to separate the wheat from the chaff :-)
Firstly, you should answer yourself that what made you choose this algorithm for your problem. i.e . the reason behind your choice.
1. The algorithm chosen should be simple and also effective.
2. If you understand some popular algorithms, then using your engineering problem, may be a production,design or industrial engineering, try these algorithms and have a comparison of solutions. Then based on the best solution obtained, you can choose one algorithm. If all algorithms are equally good in their solutions stick to the simple algorithm in terms of algorithm specific parameters. (Because tuning the parameters is a big head ache. ).
3. Finally, I suggest you see our algorithm VOMMI ( a very optimistic method of minimization) in comparison to your EWA and find the best.
All the best
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling