Different discussions but same response:

I do not recommend to attempt to list out metaheuristics according to specific names and terminologies, as most recent "nature-inspired" methods rely on artistic/fancy terminologies but boil down to a few core optimization concepts which are already known. I recommend to read this excellent position paper first:

Article Metaheuristics -- the metaphor exposed

As well as some recent surveys that analyse the core success factors of recent metaheuristics:

Article Metaheuristics in Combinatorial Optimization: Overview and C...

Article Heuristics for Multi-Attribute Vehicle Routing Problems : A ...

As a community, we should orient our effort on an analysis and unification of optimization methods under simple technical names, rather than wasting critical time (and reputation) by inventing new names and creating micro-societies dedicated to cats, birds, frogs and predating water drops.

not clear whether this provides in practice any advantage over more standard approaches of which i have found genetic algorithms to work well

Dear Colleague:

I agree with James. If you want to see all the things that inspired "new" metaheuristics, check the following "bestiary"

http://conclave.cs.tsukuba.ac.jp/research/bestiary/

Different discussions but same response:

I do not recommend to attempt to list out metaheuristics according to specific names and terminologies, as most recent "nature-inspired" methods rely on artistic/fancy terminologies but boil down to a few core optimization concepts which are already known. I recommend to read this excellent position paper first:

Article Metaheuristics -- the metaphor exposed

As well as some recent surveys that analyse the core success factors of recent metaheuristics:

Article Metaheuristics in Combinatorial Optimization: Overview and C...

Article Heuristics for Multi-Attribute Vehicle Routing Problems : A ...

As a community, we should orient our effort on an analysis and unification of optimization methods under simple technical names, rather than wasting critical time (and reputation) by inventing new names and creating micro-societies dedicated to cats, birds, frogs and predating water drops.

I agree - the sheer volume of metaheuristics makes me feel like I am at a much too large buffet, where 95% of the food and drink is already familiar, and therefore dull, and the remaining 5% of the dishes make not sense at all - meatballs with corn flakes on top, and a bowl of cereal and Coca-Cola. I won't stay long at such a place.

True ! Some Journals are beginning to cast doubts on metaheuristics as not being original enough

My experience is the following: I tried to design dozens of "abstract" metaheuristics but the best one was the only one that was nature-inspired. So it seems that nature-based inpiration has some value. I agree that many metaheuristics are found and published. However, this might not be a problem. First, finding that nature uses the same or similar algorithms in different species is a form of validation of that algorithm. Second, while many metaheuristics look similar, I found that a small difference in the algorithm can have a huge impact on the performances.

Jean-Baptiste: that is NOT the dynamics at work here; the development of metaheuristics appear to me to be built on two principles: (a) trying to adjust or combine one or two mechanisms from previous heuristics, and (b) trying to find a name that hasn't been used yet - and I wonder when the alphabet no longer is enough. In all my years of doing science, I have never before experienced a more dull practice than this. (And mark my word: the fact that a method mimics something "natural" may have very little bearing on whether it works or not.)

At the risk of repeating myself, I found the following papers thought-provoking, and fair:

Sörensen, K. (2015). Metaheuristics—the metaphor exposed. International Transactions in Operational Research, 22(1), 3-18.

Weyland, D. (2015). A critical analysis of the harmony search algorithm—How not to solve sudoku. Operations Research Perspectives, 2, 97-105.

Fister, I., Mlakar, U., & Brest, J. (2016). A new population-based nature-inspired algorithm every month: Is the current era coming to the end?

Bird, snakes, bees, etc... what is next ? someone has to definitely fix and clarify the difference and gap between exact approaches and metaheuristics.

All the answers are great

Hello

I think that the pro-metaheuristic exaggerated by proposing some ridiculous inspirations...... it's the time to make end to this.

I think that there are few metaheuristics with a solid theorem and inspiration , and the other "novel metaheuristics" are simply a renaming of the following:

genetic algorithms

particle swarm optimization

ant colony

GRASP

simple local search

simulated annealing

tabu search

VNS.

the "new" metaheuristics are like Bloatware in phones.

Most of the algorithms are based on several process that according to determined point of view can be considered search strategies. They frequently represent only marginaly the process in which is supously inspired

Now I am a bit more in providing answers again (to other threads, of course, but there they even mentioned my name and asked me to respond; I shall not do that all the time, but now it seems ok ).

Do you know the field? Here is a paper. Do you know it?

Article Semi-Greedy Heuristics: An Empirical Study

Yes - perfect. No - too bad.

Is this a metaheuristic?

Is GRASP (invented later) the same?

Ok, you shall find the answers. The questions are more than suggestive.

Let me add a bit regarding the birds. I possibly hate or love migrating birds (another of these more or less new metaphors). But rather than just discussing here what I usually do is exploring, researching, and even publishing. Then I have a foundation for discussing. Isn't that a good way? I think it is. Kindly have a look:

Article On the performance of the hybridisation between migrating bi...

Conference Paper Analysing the Performance of Migrating Birds Optimisation Ap...

And there is one more (less birdie, though):

Conference Paper Hybridisation of Evolutionary Algorithms Through Hyper-heuri...

That was related to a competition. After the conference we received some referee reports. I liked the one most which said that our approach is too simple and should possibly be rejected, but "Unfortunately, they won the competition."

Did I provide an answer? Did I provide some evidence for something? No - just food for thought.

:-) Great response. Indeed, we cannot know what works in the world of combinatorial/integer/global optimization; our hope is that we see something familiar in one or more representative models, and with the experience we have - and the existing literature - we might design a few tentative schemes for the problem at hand. Luckily I am in academia - I can be nerdy when I want to, without any potential negative consequences at the other end. But I can't do this too much: yesterday I got a manuscript accepted after 4 hours of refereeing by the womderfully cranky Journal of Irreproducible Results, so that was probably my only shot this year. Even us academicians need to let off steam, and that is one way: the use of humour/comedy - we are of course very well aligned with the old comedians, or so I often feel among my productive and very serious engineering fellows.