Ant colony optimization algorithm

Поділитися
Вставка
  • Опубліковано 18 тра 2023
  • Ant colony optimization algorithms (ACO) is a probabilistic technique for solving computational problems which can be reduced to finding good paths through graphs. Artificial ants stand for multi-agent methods inspired by the behavior of real ants.
    In the video, I tried to clearly explain how the algorithm works.
    You can support the channel on Patreon:
    / simulifehub
    More support, more opportunities for new projects.
    ko-fi.com/simulifehub
  • Наука та технологія

КОМЕНТАРІ • 70

  • @LaplacescherDaemon
    @LaplacescherDaemon Рік тому +25

    I really like that in this Video not only the ant or the math part is shown, but instead both

  • @LoisoPondohva
    @LoisoPondohva Рік тому +35

    Eh, liked the last voice better
    But other than that, great video

    • @btdubsy
      @btdubsy Рік тому +6

      the other voice is 50% of why i watch

    • @jimster1111
      @jimster1111 Рік тому

      wasnt the other voice real? had a european accent and everything.

    • @LoisoPondohva
      @LoisoPondohva Рік тому +5

      @@jimster1111 the last one sounded like an AI too for me, just a better one. But might've been a narrator, doesn't matter too much to me, what matters is that this one is hard to listen to.

    • @tisajokt7676
      @tisajokt7676 Рік тому +1

      @@LoisoPondohva a lot harder to listen to, yeah

  • @hisao1291
    @hisao1291 Рік тому +7

    This was an amazing video to explain the concept. Thanks for making it this complete !

  • @dimitri0404
    @dimitri0404 Рік тому +80

    Can we make a subreddit or discord? It would be nice for people who watch this channel to share their own attempts at implemanthing these algorithems.

    • @lettuceturnipthebeets790
      @lettuceturnipthebeets790 Рік тому +7

      the author is quite busy with his current work and studying for data science, also translating the videos takes some space, so I doubt he'd have spare time managing a discord channel or a subreddit :

    • @tobenamed610
      @tobenamed610 Рік тому +3

      @@lettuceturnipthebeets790he actually does have a discord for patrons only, it’s very dead right now though.

    • @leviearle4623
      @leviearle4623 Рік тому +2

      I doubt I would be doing any implementation, but that would be cool to look in on and see multiple peoples attempts/tests

    • @krish4659
      @krish4659 7 місяців тому

      @@leviearle4623 do you any thing how the 16:01 global pheromones are updated , please help me up

  • @relativepi5046
    @relativepi5046 Рік тому +2

    This is one of the coolest things I have seen, and I love how you explained it!

  • @looalx_WD
    @looalx_WD 9 місяців тому +1

    CRAZY GOOD QUALITY in terms of clarity and knowledge. THX MUCH

  • @crusatyr1452
    @crusatyr1452 Рік тому +13

    You saying "vertice" instead of "vertex" took a second for me to get used to lol

    • @MyDude199
      @MyDude199 Рік тому +9

      They use an ai voice cause im pretty sure they are a non-native english speaker

    • @crusatyr1452
      @crusatyr1452 Рік тому +2

      ​@@MyDude199 No yeah, but that's not entirely relevant because it's a common alternative to "vertex", at least common enough to be on Wiktionary.

    • @josiahmanson
      @josiahmanson Рік тому +3

      @@crusatyr1452 being a non-native english speaker is relevant. a native speaker would know that vertex doesn't follow the regular rule for pluralization of just adding an 's'. typing vertice gets a red squiggle from the spell checker while typing this comment, but my spell checker does not suggest the correct singular form, so i could easily see that even if it was marked as wrong, it isn't clear how to correctly fix.

    • @crusatyr1452
      @crusatyr1452 Рік тому +1

      @@josiahmanson
      No, I completely agree.
      My reply wasn't the most precise way of conveying what I meant to, admittedly. My main point was that it's common enough among native-speakers to have a place on Wiktionary (if you except that premise).
      By "relevant" I could've said smth like "necessary" in the sense that the error alone doesn't necessary mean that they're non-native.

    • @nester3190
      @nester3190 Рік тому +1

      ​@@crusatyr1452 In the channel description the author clearly states that this is the English version of the Russian foo52ru channel

  • @abironnoy3115
    @abironnoy3115 Рік тому +3

    Thank you so much. The animations were very helpful to understand the whole concept specially the iteration part

  • @pfever
    @pfever 2 місяці тому

    great video! I really appreciate how every formula is explained in detail!

  • @meanpillscasper
    @meanpillscasper Рік тому

    This video was crystal clear. Thank you very much sir.

  • @kylemorris5338
    @kylemorris5338 Рік тому +16

    Solving a problem via algorithm whose brute-force solution would require "all computers and billions of years" is pretty much the textbook example of "work smarter, not harder"

    • @Yay295
      @Yay295 Рік тому +3

      This algorithm doesn't necessarily find the best result though.

    • @lettuceturnipthebeets790
      @lettuceturnipthebeets790 Рік тому +8

      @@Yay295 as they say, best is enemy of good enough :>

    • @Yay295
      @Yay295 Рік тому +1

      @@lettuceturnipthebeets790 Sure, but usually when people say "work smarter, not harder", the expectation is that the end result is the same. This algorithm isn't that much smarter than brute-force - you're still comparing complete solutions against each other to see which is best - this algorithm is only faster because you're not checking all of the solutions. If you actually did check all of the solutions using this algorithm it would actually be slower than brute-force because of the overhead of the algorithm itself.

    • @thinkandmove479
      @thinkandmove479 28 днів тому

      Yeah, then work smarter and not harder, and prove that P != NP

  • @VijayVighnesh
    @VijayVighnesh Рік тому

    A good video to get clarity on ANT colony optimization...

  • @fabricehategekimana5350
    @fabricehategekimana5350 8 місяців тому

    Thanks for this instructive video !

  • @b10ckh0usesurv10rs
    @b10ckh0usesurv10rs Рік тому +2

    I think that the first voice was better, also a suggestion from me is that if you made one of those simulation games with this and added the cumulative knowledge of previous videos you could become more popular

  • @shauryasaxena5263
    @shauryasaxena5263 4 місяці тому

    Thanks a lot sir for such a nice suggestion

  • @Xx_babanne_avcisi27_xX
    @Xx_babanne_avcisi27_xX Рік тому +13

    last voice better

    • @TheBlindfischLP
      @TheBlindfischLP Рік тому +2

      Definitely, I didn't even notice at first, that the voice in the last video wasn't a human.

    • @dustinmock5016
      @dustinmock5016 Рік тому +2

      I agree, the old one also reminds me of someone smart, this one is just like a generic AI voice

    • @Xx_babanne_avcisi27_xX
      @Xx_babanne_avcisi27_xX Рік тому +2

      @@dustinmock5016 bro i agree with that sooo much.

  • @tar0nja
    @tar0nja Рік тому +2

    Honey wake up, Simulife Hub posted a new video!!!

  • @Paladiea
    @Paladiea Рік тому +3

    Vertex is the singular of vertices.

  • @qu765
    @qu765 Рік тому +4

    Could you make a video about algorithm that optimize 2d graphs.
    So a graph with say 200 nodes where each node has 1-3 connections to other nodes and it needs to move the nodes in a way that is easy to see the connections by a human.
    I had this problem (and still do) for a hobby of mine and was only able to make an algorithm that worked up to about 100 nodes before manually manipulating it results in a better graph.
    I am wondering if you could have better results.
    The algorithm that I use a physic simulation kinda. Every node has velocity and moves with it (although there is a lot of friction) and then they get a force towards connected nodes and a force away from other ones.

  • @revimfadli4666
    @revimfadli4666 Рік тому +2

    I wonder if this would be effective for the economic simulation mentioned in the previous video

  • @nidhalabidi69
    @nidhalabidi69 3 місяці тому

    Amazing explanation! I have a question: When calculating the probability of moving from city_i to city_j at 11:58, what should we do when the amount of pheromone is 0 on all edges. According to the formula we're going to divide by zero!! As a solution for this can we give it a small non-zero initialization?

  • @abelvictor8322
    @abelvictor8322 Рік тому

    What do the parameters alpha and beta really do ? Like why do you wannmt to change them ?

  • @SanyaBossDota
    @SanyaBossDota Рік тому

    Здравствуйте. Можете пожалуйста выложить исходный код из видео Нейронная сеть. Часть 7. Рисунок штрихами из фотографии. Очень буду благодарен.

  • @kamalibraimm2751
    @kamalibraimm2751 5 місяців тому

    How can I implement a swarm metaheuristic on cvrp

  • @user-ml9ln7mo7z
    @user-ml9ln7mo7z Рік тому

    me vibing with that ant in the beginning

  • @krish4659
    @krish4659 7 місяців тому

    how is global pheromone calculated for this 16:01
    please help me T_T i will recommend video to my friend if you can clarify this doubt T_T i reallly need it

  • @samajhseva5909
    @samajhseva5909 Рік тому

    I need to connect with you how can i please tell

  • @dustinmock5016
    @dustinmock5016 Рік тому +2

    Please go back to using the old text-to-speech voice, I liked that one a lot more

  • @THE_ONLY_REAL_WAFFLE
    @THE_ONLY_REAL_WAFFLE Рік тому +1

    The ant dancing at the start be like:
    (⁠☉ ☉⁠ )
    (⁠ ☉ ☉⁠ )
    (⁠ ☉ ☉)

  • @kadhiravensamynathan6001
    @kadhiravensamynathan6001 7 місяців тому

    It's 3am and idk what I'm doing here. I've no idea about the formulas or ants.

  • @user-sw3tf6jw5z
    @user-sw3tf6jw5z Рік тому +1

    Here's an original video
    ua-cam.com/video/8KTzAiusfPs/v-deo.html

    • @wallcraft-video
      @wallcraft-video  Рік тому +2

      Both channels are mine. :)
      Decided to expand the audience

  • @JT.CAWTHRON
    @JT.CAWTHRON Рік тому

    I prefer the previous voice

  • @michaelrotchell6164
    @michaelrotchell6164 10 місяців тому +1

    Like number 777

  • @user-vx7lg5wm5l
    @user-vx7lg5wm5l Місяць тому

    Korean sub plz~~😥

  • @user-mj4nt1om2k
    @user-mj4nt1om2k Рік тому +1

    second

  • @_Gusza
    @_Gusza Рік тому +1

    first, ig

  • @adorman
    @adorman Рік тому

    Please change the voice back.

  • @KSPL9999
    @KSPL9999 Рік тому +7

    That voice is to annoying

    • @jaimalad2902
      @jaimalad2902 Рік тому

      I think its AI generated voice

    • @KSPL9999
      @KSPL9999 Рік тому

      @@jaimalad2902 of course it is

    • @LaplacescherDaemon
      @LaplacescherDaemon Рік тому

      I think it is not distracting at all, for me there is no problem with using an AI-Voice. If he dosn't like to present his real voice to us, he shouldn't

  • @jaimalad2902
    @jaimalad2902 Рік тому +1

    This video is AI generated, voice actually is not real voice. WTH! Content is somewhat OK