The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.

The result was keep. Although there is a merge !vote, the keep !votes based on policies outweigh it. (non-admin closure) Anarchyte (work | talk) 13:56, 29 July 2016 (UTC)[reply]

Firefly algorithm[edit]

(Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)

This article is part of the following group of articles that I have all nomination for deletion (individually):

These article all detail research done by Xin-She Yang. All suffer from the following problems:

Most citations include Yang as one of the authors (i.e. are primary).
Citations numbers of the article look superficially impressive, but include many self-citations and even reek of a citation circle.
Articles have been created by Category:Suspected Wikipedia sockpuppets of Metafun, why likely is Yang himself.
I could not find any respectable overview books and articles describing this work as considered relevant in the field. —Ruud 14:25, 15 July 2016 (UTC)[reply]

5940 papers

The Firefly algorithm is discussed in many chapters of the edited book Adaptation and Hybridization in Computational Intelligence [1]. Although the first chapter is indeed written by Yang, the other chapters are not. You can see the list of chapters and authors from Amazon's "Look Inside".[2] Also, on Google Scholar there are 5,940 search results for "firefly algorithm", of which only a dozen or two are authored by Yang. [3] I recommend the Refimprove template be added rather than the article be deleted. Michaelmalak (talk) 16:05, 15 July 2016 (UTC)[reply]

Yes, as I noted a superficial look makes this look quantitatively well-cited, but superficial looks can be deceptive. From all those references could you pick out those which you believe have been published in the qualitatively best venues? Did you find any truly respected textbook or overview article discussing this? —Ruud 20:23, 15 July 2016 (UTC)[reply]

Searching Google Books for "firefly intitle:optimization" [4] turns up 511 books from various well-known publishers, including:

* Evolutionary Optimization Algorithms By Dan Simon, Wiley, [5]

* Computational Intelligence Paradigms for Optimization Problems Using MATLAB®/SIMULINK® by S. Sumathi, L. Ashok Kumar, Surekha. P, CRC Press [6]

* Encyclopedia of Business Analytics and Optimization, edited by Wang, John, IGI Global [7]

Michaelmalak (talk) 03:34, 16 July 2016 (UTC)[reply]

Evolutionary Optimization Algorithms looks like a decent enough source. Are there one or two more like that? (The other two you suggested above are already seem to be of more dubious quality. I couldn't find out anything about the authors of the CRC Press one. There is another similarly titled book, though. Neither appear to have received any citations. Somewhat similar story about the IGI Global one, which isn't a top-tier publisher to begin with.) —Ruud 09:34, 16 July 2016 (UTC)[reply]

Weyland, Glover, Sörensen

Weyland, who has previously criticized another "nature-inspired" metaheuristic (harmony search), also explicitly calls out the firefly algorithm as being of unclear novelty in the introduction of his new article (Elsevier ScienceDirect link). —Ruud 09:56, 16 July 2016 (UTC)[reply]

In recent years a huge number of novel metaheuristics were proposed. These metaheuristics are usually based on metaphors describing natural processes or social phenomena. The metaphors used to derive the working mechanisms of such novel metaheuris- tics are getting increasingly absurd and the connection between the metaphors on the one hand and optimization on the other hand is getting increasingly vague. It is not really clear what the flow of water [1], the leaps of frogs [2] or a salmon run [3] have to do with optimization. Additionally, it seems that the underlying working mechanisms of these methods are very similar, and in some cases even identical, to those of well-established heuristics. For example, the differences between the particle swarm optimization meta- heuristic [4] and ‘‘novel’’ metaheuristics like the firefly algorithm [5], the fruit fly optimization algorithm [6], the fish swarm opti- mization algorithm [7] or the cat swarm optimization algorithm [8] seem negligible. Nevertheless, the literature is full of results which certify exceptional performance to these ‘‘novel’’ methods. Obviously, there is something going wrong. This whole develop- ment had been ignored for quite a while, but recently open crit- icism has emerged.

Glover and Sörensen also comment on the problem we are seeing here in their Scholarpedia article under the section "The metaphor controversy":

A large (and increasing) number of publications focuses on the development of (supposedly) new metaheuristic frameworks based on metaphors. The list of natural or man-made processes that has been used as the basis for a metaheuristic framework now includes such diverse processes as bacterial foraging, river formation, biogeography, musicians playing together, electromagnetism, gravity, colonization by an empire, mine blasts, league championships, clouds, and so forth. An important subcategory is found in metaheuristics based on animal behavior. Ants, bees, bats, wolves, cats, fireflies, eagles, vultures, dolphins, frogs, salmon, vultures, termites, flies, and many others, have all been used to inspire a "novel" metaheuristic. A more complete list can be found in Fister et al (2013). As a general rule, publication of papers on metaphor-based metaheuristics has been limited to second-tier journals and conferences, but some recent exceptions to this rule can be found. Sörensen (2013) states that research in this direction is fundamentally flawed. Most importantly, the author contends that the novelty of the underlying metaphor does not automatically render the resulting framework "novel".

Given the big controversy and huge number of book and paper citations, I would think it makes it all the more important to keep in Wikipedia, and add in the criticism (with cites) -- including in the lead. Michaelmalak (talk) 12:30, 16 July 2016 (UTC)[reply]

Keep. Verifying almost 60 references is a daunting task for anyone. Many of these references seem to be conference proceedings. Although these are sometimes more "significant" in CS than perhaps in other areas of science, typically conference proceedings are not secondary sources, as required by WP:NOR. Focusing on more traditional print media, some of the remaining references appear to be from open access publishers. For example, the "Journal of Applied Mathematics", "Int. J. Computer Science Issues". Several of the other journals seem otherwise dodgy. But there is a solid core here of high-quality journals, for example the Proceedings of the Royal Society. I would really like to see the references pruned to some more manageable collection of high-quality sources. I do not think specific implementation details belong here. Entire sections are sourced only to conference proceedings, and should be removed. Also, given the amount of skepticism that these metaheuristics have generated, I find the tone of the article inconsistent with WP:NPOV, and especially WP:WEIGHT. Possibly WP:FRINGE is relevant. However, AfD is not cleanup. Although what we have right now is a pretty terrible article, it does not rise to the level of TNT. Sławomir Biały (talk) 13:16, 16 July 2016 (UTC)[reply]

Note: This debate has been included in the list of Mathematics-related deletion discussions. Shawn in Montreal (talk) 14:28, 16 July 2016 (UTC)[reply]

Keep. My general belief is that much of the work in this sort of metaheuristic is junk science, but the high citation counts and numbers of hits for this topic in Google Scholar make clear that, regardless of that, it is notable. —David Eppstein (talk) 17:22, 16 July 2016 (UTC)[reply]

2016-07-17 Vast Improvement in Article

Thank you Ruud for your editing work on this article: adding the criticism and deleting the fluff. And thank you for leaving in the actual pseudocode as that answers the question pertinent to Wikipedia readers: "What is the Firefly Algorithm"? Michaelmalak (talk) 12:58, 17 July 2016 (UTC)[reply]

Nature-Inspired

A direct quote from Xin-She Yang himself from his book Nature-Inspired Metaheuristics which has been repeatedly published by Elsevier.

Researchers have drawn various inspirations to develop a diverse range of algorithms with different degrees of success. Such diversity and success do not mean that we should focus on developing more algorithms for the sake of algorithm developments, or even worse, for the sake of publication. We do not encourage readers to develop new algorithms such as grass, tree, tiger, penguin, snow, sky, ocean, or Hobbit algorithms. These new algorithms may only provide distractions from the solution of really challenging and truly important problems in optimization. New algorithms may be developed only if they provide truly novel ideas and really efficient techniques to solve challenging problems that are not solved by existing algorithms and methods.

Sadly, the scientific community rewards those algorithms that are able to produce better results on a set of benchmark functions Test functions for optimization . Coincidentally, these "inspired algorithms" have been performing well in solving such test cases along with other complex problems, hence the high number of citations. Although, these algorithms may appear to be "metaphoric", most of the original algorithms in this field share at some level, the same level of similarity in terms of "population", "fitness", "operators", "solutions" etc. Hence, singling out "inspired" algorithms for deletion based on a few handful of publications outlining its negative "novelty" against the large number of publications outlining its "effectiveness" is still a matter up for debate. It is true that research at this point of time is mired at the metaheuristic level but till the time the scientific community decides over the debate of "fittest" vs "novelty" , as an knowledge sharing site, both the pros and cons should be weighed infront of the reader, meaning both the applications that have been conspicuously blanked for some algorithms due for deletion and the criticism like the one already been put for firefly should together be put up as information. Furthermore, to clarify some of the claims but these "algorithms" have been published not only in 2nd tier journals or conferences but reputed journals like Elsevier , Springer Publishing , Institute of Electrical and Electronics Engineers, wiley etc. Capn Swing (talk) 12:02, 19 July 2016 (UTC)[reply]

Keep as there's enough to suggest it is convincing to keep. SwisterTwister talk 19:08, 25 July 2016 (UTC)[reply]
Merge People here are making pretty strong cases for deletion by citing these damning criticisms, and then saying "keep" for some reason. The problem here is that the articles/algorithms themselves aren't notable, but the "Nature-inspired metaheuristics" category is notable (for the criticism if nothing else). Rather than keep them all and add a "criticism" section to every one, why don't we make a succinct article for all of them? I'd also like to point out that being mentioned and being notable are not the same thing. A bunch of academics copying and adding to lists en-masse can quickly lead to citation bloat, as we've seen in the WP article itself. These topics should be treated the way the academic world treats them: as an example mentioned in passing by people who don't care how (or if) they work. Jergling (talk) 19:53, 25 July 2016 (UTC)[reply]

I do think the number of metaheuristic algorithm articles needs to be reduced, but I believe Firefly and Artificial Bee Colony are notable enough for articles of their own. Michaelmalak (talk) 21:27, 25 July 2016 (UTC)[reply]

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.