Talk:Swarm intelligence

Template:IEP assignment

WP:SYNTH

Ther is a bunch of algorithm examples here. Please provide references which describe them as SWINT otherwise this constitutes WP:SYNTH. -M.Altenmann >t

River formation dynamics

This thread had been moved here from a user talk page, since this is a more appropriate place. The thread was preceded by the removal in the article of the section called River Formation Dynamics.

---

You asked: "(where is the rationale on the talk page?) (undo | thank)" - the talk page is extremely small. And the rationale is right there: WP:SYNTH. It is your job to provide references which explicitly say that rivers are swarm intelligence, not mine to engage with you in theoretical discussions about AI. -M.Altenmann >t 08:35, 5 March 2015 (UTC)

You were the one who deleted the section on the grounds that it was "irrelevant". In fact, casual inspection of the seminal paper shows clearly that it ticks all the boxes required to be classified as swarm intelligence. That's not synthesis. I don't see you deleting the other algorithms on the the grounds that they need such references. However some references which explicitly tell you that River Formation Dynamics is a form of swarm intelligence include:

• Mishra, E. A., Das, M. N., & Panda, T. C. (2013). Swarm Intelligence Optimization: Editorial Survey. International Journal of Emerging Technology and Advanced Engineering, 3(1). Full text – see page 219.
• Goel, L., Gupta, D., Panchal, V. K., & Abraham, A. (2012, November). Taxonomy of nature inspired computational intelligence: A remote sensing perspective. In Nature and Biologically Inspired Computing (NaBIC), 2012 Fourth World Congress on (pp. 200-206). IEEE. Full text – see pages 201–202.
• Krause, J., Cordeiro, J., Parpinelli, R. S., & Lopes, H. S. (2013). A survey of swarm algorithms applied to discrete optimization problems. Swarm Intelligence and Bio-inspired Computation: Theory and Applications. Elsevier Science & Technology Books, 169-191. Full text – see pages 1–2.
• Binitha, S., & Sathya, S. S. (2012). A survey of bio inspired optimization algorithms. International Journal of Soft Computing and Engineering, 2(2), 137-151. [Full text – see page 142.
• Mahant, M., Choudhary, B., Kesharwani, A., & Rathore, K. S. (2012). A profound survey on swarm intelligence. International Journal of Advanced Computer Research, 2(1), 31-36. Full text – see page 35.

There are many others. I have reinstated the section. --Epipelagic (talk) 10:18, 5 March 2015 (UTC)

Thank you for collecting these references. Now, the next step is to add 1-2 of them to the section with the words: "this method is classified as SwInt". While for ants it is evident, it is not at all for river, because there is no direct analogy. -M.Altenmann >t 15:46, 5 March 2015 (UTC)

By the way, some of your refs contain good summaries/surveys of the method. I assume you be interested in the domain. Why don't you improve this really poor article basing on the findings above? -M.Altenmann >t 15:52, 5 March 2015 (UTC)

Gravitational search algorithm

@Altenmann: I see you have now deleted the gravitational search algorithm as well, claiming there is "nothing to discuss". Here are some recent reviews on swarm intelligence that do not agree with you: [1] [2] [3] [4] [5]. Please disengage from editing this article unless you are willing to familiarise yourself somewhat with the topic. --Epipelagic (talk) 04:32, 29 August 2015 (UTC)

I was claiming "nothing to discuss" because we discuss only references, not wikipedian's opinions. Your references do not describe it as swarm intelligence. So here is my advice back to you: please disengage from editing wikipedia unless you are willing familiarize yourself with rules of providing references. In particular, I fail to understand why you are putting these links into talk page instead of article. YOu don't have to convince me personally; you have to make the article verifiable. - üser:Altenmann >t 16:18, 29 August 2015 (UTC)

You say the references I gave you "do not describe it as swarm intelligence". Let me hold your hand again and walk you through your issue. Look at figure 2 on page 201 of [6]. Now look at table 1 on page 219 of [7]. Then read the second paragraph of the introduction to [8]. Finally ponder the section on GSA (gravitational search algorithm) on pp. 34–35 of [9]. --Epipelagic (talk) 19:54, 29 August 2015 (UTC)

Sorry, I was not clear. Your links in the talk page are not references. The references are in the article, In the talk page it is just chat. The reference in the article to the authors of the algorithm says:

In recent years, various heuristic optimization methods have been developed. Many of these methods are inspired by swarm behaviors in nature. In this paper, a new optimization algorithm based on the law of gravity and mass interactions is introduced. In the proposed algorithm, the searcher agents are a collection of masses which interact with each other based on the Newtonian gravity and the laws of motion.

The authors clearly say that their method imitates law of gravity, not swarm. That future fuzzy-minded secondary sources decided that gravity is swarm, it is a different story. And the latter statement is not referenced in the article. While I find it dubious, if someone classified it so, then so be it. But it must be referenced in the article; I don't need you holding my hand, just fix the damn article. - üser:Altenmann >t 21:02, 30 August 2015 (UTC)

That's truly priceless Altenmann. You should offer the clarity of your services to the "fuzzy-minded" swarm intelligence community. But please now give me a helping hand. Specifically where did the "authors clearly say that their method imitates law of gravity, not swarm"? Of course it needs to imitate gravity because otherwise it wouldn't be a gravitational algorithm. But where, as you claim, do they also clearly say it doesn't "imitate" a swarm? --Epipelagic (talk) 06:11, 31 August 2015 (UTC)

OK, colleague we exchanged a bit of irony towards each other, please accept my apologies for taking a wrong bait. Now to the real issue and none of personal stuff (even towards anonymous SwInt community). Per wikipedia rules, it is the job of the wikipedia who makes a statement to provide evidence for the statement. In our case, the article implicitly assumes that "Gravity Algorithm" is of "swarm" kind. The article cites the papers of GrAlg originators. I say that the originators seem to not explicitly state that theirs are swarm algorithm, so I ask to provide a citation from respectable AI sources to this end. Therefore please not be offended by my original tone and address the issue per wikipedia rules of WP:CITE. Obviously I could have just copied/pasted your links into the article, but I am leaving this to you because IMO you are in a better position to put them into a proper place (and not because I am a stubborn nitpicker). - üser:Altenmann >t 15:35, 31 August 2015 (UTC)

That said, I can understand where it comes from. I understand that the term "swarm" in this context should not be taken literally, that it is a mere analogy used to give a good name for a certain category of algorithms. And the article must have clearly explained that; of course, citing someone from SwInt community, not in the form of original research akin to my previous sentence. - üser:Altenmann >t 15:35, 31 August 2015 (UTC)

Attempt to bias the discussion

User Ruud Koot according to his profile "a Ph.D. student in computer science" has made several changes to pages related to the Swarm Intelligence area. He removed a lot of material, and numerous references and links from many Wikipedia pages on metaheuristics. He removed nearly all sub-sections on nature-inspired metaheuristics from the Swarm intelligence page, and grouped them in a new page List of metaphor-based metaheuristics, after a long section on “Criticism of the metaphor methodology”. According to his view, nature-inspired metaheuristics hide “their lack of novelty behind an elaborate metaphor”. He also copied and pasted the following statement at the beginning of many metaheuristics pages.

“Nature-inspired metaheuristics in general have started to attract criticism in the research community for hiding their lack of novelty behind an elaborate metaphor.[5][6][7][8][9] In response, Springer's Journal of Heuristics has updated their editorial policy to state that:[10] Implementations should be explained by employing standard optimization terminology, where a solution is called a "solution" and not something else related to some obscure metaphor (e.g., harmony, flies, bats, countries, etc.).”

Amongst the pages affected by his changes there are those related to the Bees algorithm, Artificial bee colony algorithm, Harmony search, Cuckoo search, Glowworm swarm optimization, and Firefly algorithm (I suspect many more).

Whilst I am not against scientific debate and criticism, I believe that Koot’s pervasive and arbitrary changes are a clear attempt to bias the debate according to his views. The statement about Nature-inspired metaheuristics is at best disrespectful, and ignorant of the impact of such techniques (over 6000 citations on the topic only this year according to WoS). His re-editing of the Swarm Intelligence page shows a very partial view of the subject, and discarded some of best known and widely used metaheuristics in the field. For example, the first article on the Bees Algorithm (Pham et al., 2006) was cited so far 927 times, the first article about Artificial Bee Colony (Karaboga, 2005) was cited 3090 times, Cuckoo search (Yang 2009) 1652 times, Firefly algorithm (Yang 2009) 1248 times. There is no way to undo the changes, as he promptly rolls back his version. He semi-protected the ‘Swarm Intelligence’ page to protect his changes. I hereby ask him to roll back his changes, on this page and related ones. I have no objection about him adding an extra section with his personal views about the algorithms, but I believe he should respect the contributions of many experts in the field. Marco castellani 1965 (talk) 00:58, 25 August 2016 (UTC)

To summarize your complaint: Several algorithms have been removed from the section Algorithms. I'd like to give the following justification for this:

• First of all, most of these algorithms have not been completely removed from Wikipedia, but instead have been moved into the new article List of metaphor-based metaheuristics. Encyclopedic articles are intended to give broad overviews on particular topics, not exhaustively list every minutia (this is better left to list articles.) As noted in 'Lones, Michael A. (2014). "Metaheuristics in Nature-Inspired Algorithms" (PDF). GECCO '14. doi:10.1145/2598394.2609841.':

Evolutionary algorithms (EA), particle swarm optimization (PSO), ant colony optimization (ACO) and their variants dominate the field of nature-inspired metaheuristics.

As an overview article it would thus be best to list these as representative examples of swarm intelligence algorithms. Also see the related policies WP:UNDUE and WP:NOT.

• As noted in the sources cited in List of metaphor-based metaheuristics § Criticism, including:
  • Weyland, Dennis (2010). "A Rigorous Analysis of the Harmony Search Algorithm: How the Research Community can be Misled by a "Novel" Methodology" (PDF). International Journal of Applied Metaheuristic Computing. 1 (2): 50–60. doi:10.4018/jamc.2010040104.
  • Sörensen, Kenneth (2013). "Metaheuristics—the metaphor exposed". International Transactions in Operational Research. 22: 3–18. doi:10.1111/itor.12001.

the "novelty" of many of the algorithms published since the originals (EA, ACO, PSO, ...) has been strongly put into question. If we want to give a balanced overview, we cannot simply list the newer algorithms without these critical notes. Such an extended discussion would go a bit too far abroad for this article and is better left to List of metaphor-based metaheuristics. Also see the related policies WP:UNDUE and WP:NPOV.

—Ruud 21:03, 25 August 2016 (UTC)

You have made extensive changes to several Wikipedia pages concerning the general Swarm Intelligence topic, removing a lot of scholarly material and presenting the topic under your very personal slant. You seem to over-rely on a very selective list of sources, which you have copied and pasted all over Wikipedia. The scholarly debate on the usefulness, originality, and merits of metaheuristics is at least 30 years old; you have chosen one very narrow stance and biased the entire presentation of the field on Wikipedia. You have removed a lot of information and references (e.g. applications of the Bees Algorithm to real world complex problems) which give the reader examples of the usefulness of such techniques, and replaced this information with your own personal negative view of the topic. Whilst I don't object to your freedom to express your views in a respectful manner, I do object to the fact that you biased the entire presentation on Wikipedia of a very large and complex topic. I also do object to the fact that you removed a large amount of contributions of many other authors, effectively vandalising several pages. Whilst I don't object to your idea of moving to a dedicated page the list of nature-inspired metaheuristics, I do object to the way you presented these algorithms, preceding them with a long critical text copied and pasted from your favourite authors. Again, based on your very personal and negative point of view, you have presented the whole field of nature-inspired metaheuristics under a very negative slant. What you have left on the Swarm Intelligence page is also not a sound representation of the field. Differential Evolution does not belong to the Swarm Intelligence family, but is one implementation of Evolutionary Algorithms. Clonal selection belongs to the family of Artificial Immune Systems. There are fundamental differences between these classes of algorithms (i.e. decentralisation), and I don't think they should be mixed together. From the point of view of accuracy, the page was far more sound before your modifications. Before it presented 'Swarm Intelligence' as 'nature-inspired metaheuristics' (improper but widely used), now it is a jumble of heterogeneous algorithms. In any case, I do not want to engage with your personal ideas on the subject. I question your behaviour. When one user removes a significant portion of material, links, and references on several Wikipedia pages to push forward his personal negative view of a topic, I believe we have a 'democracy' problem. From your profile, I understand you are a doctoral student in a field which is not even related to the topic you have hijacked. In this case, I believe we have also an authoritativeness problem. Whilst I am aware of the fact that Wikipedia is based on a horizontal structure where all contributions are valuable, I believe you should exercise some prudence before criticising respected professors for "hiding their lack of novelty behind an elaborate metaphor". You can append your views at the end of the topics, possibly in a more succinct and respectful fashion, without removing the contributions of other authors. Therefore, for the sake of neutrality and objectivity, I ask you once again to restore the old pages, eventually adding your views in a separate section at the end.Marco castellani 1965 (talk) 00:05, 26 August 2016 (UTC)

I agree with what you said about DE and AIS/CS. They are only here because they were already listed in the article before I cleaned it up. I have not yet had the chance to move them elsewhere, but will do so eventually. I disagree with most everything else you said. —Ruud 00:21, 26 August 2016 (UTC)

Regardless of whether you agree, I ask you to restore the material you have removed from the various pages, and present the topic under a fair and impartial point of view. Marco castellani 1965 (talk) 00:47, 26 August 2016 (UTC)

The material is currently presented in a fair an impartial manner. —Ruud 00:53, 26 August 2016 (UTC)

NPOV discussion

I agree that the new changes have biased the discussion completely. This page has been improved in the past by a lot of contributors and used to reflect a completely neutral perspective. The fact that the user Ruud Koot has made the article protected shows that he is well-aware of his biased and inappropriate changes in this page as the main page of swarm intelligence. The user Marco castellani has made valid objections about your changes and I highly support him. This does not mean that I do not agree that there are metaphor-based algorithm and even worse artificially made metaheuristics. I believe that your ignorance of a neutral point of view and contributions of contributors in the past is not appropriate. Please restore the old pages and write your point of view as a new section.