How to Solve Travelling Salesman Problems - TSP

Поделиться
HTML-код
  • Опубликовано: 30 дек 2024

Комментарии • 98

  • @willpowelluk
    @willpowelluk 5 лет назад +9

    Don't normally comment on videos but thought the explanation was super clear. Thank you!

  • @jameslockwood217
    @jameslockwood217 10 лет назад

    I can happily say that this is a maths (or math) teacher at my school and I can get help like this anytime I need it!

  • @stefan4207
    @stefan4207 7 лет назад +50

    0:00 "konitshiwawa wannano"
    *Instant Skip*

  • @FlyBoyGrounded
    @FlyBoyGrounded 11 лет назад +3

    If you'd taken a different vertex out to get your lower bound, could you have found a different (possibly greater) lower bound? Also, surely it's possible, by coincidence/chance, to obtain the optimal solution when finding the upper bound if you happen upon the right starting point?

  • @alvesdds
    @alvesdds 10 лет назад +8

    You did not mention the fact that you only cann visit a point once. Are you considering that in your problem?

  • @Tokinjester
    @Tokinjester 8 лет назад +1

    can't agree with the second exapmle. the problem states the starting point is A but you start from B, so you'd have to add 16 to your total to get from A to B...no?

  • @amandaniess7028
    @amandaniess7028 7 лет назад +4

    I am writing a math project about TSP, and I am looking for information about Christofides' algorithm. Have you made any videos about that? :)

  • @alexpopov4464
    @alexpopov4464 6 лет назад +3

    In Travelling Salesman problem you have to visit vertices exactly once. And in your example you visit D twice... How is this relevant to Travelling Salesman problem?

    • @sirxavior1583
      @sirxavior1583 5 лет назад

      If I'm not mistaken all he did was run a verification algorithm. You can create/run a verification algorithm on an NP-complete problem to demonstrate that the problem is polynomial time but that doesn't mean you can create a solution to the problem in polynomial time. I could demonstrate that I could be anywhere in the world in constant time O(1) by teleportation and going directly to my destination but that doesn't mean that there exists a solution in O(1) time constant time.I hope that makes sense.

  • @michelleparker3985
    @michelleparker3985 9 лет назад

    Thank you your heart is in the right place! TSP is tricky no matter who you are!!!

  • @aoyukialquen2835
    @aoyukialquen2835 11 месяцев назад

    i somehow manage to find upper bound 78 from A > B > C > E > D > A. did i do something wrong? or what did i misunderstood?

  • @kellyredds7292
    @kellyredds7292 4 месяца назад

    Why did you start from B in your minimum spanning tree?

  • @LiaquathHassan
    @LiaquathHassan 9 лет назад +55

    You are travelling D twice, that doesn't fulfill the requirements of TSP!

  • @ramanbhardwaj408
    @ramanbhardwaj408 4 года назад

    How can in a triangle sum of two sides is smaller than the third... In triangle AEB 16+4

  • @stavone12
    @stavone12 9 лет назад +55

    You did not solve anything... Why the title?

  • @carabouzouklis
    @carabouzouklis 6 лет назад +1

    I don't think people understand the tutorial at all. He didn't find a solution because there is not always an optimal solution to the problem. What he told you is that if you find a root that has weight between those two values then it is a really good solution.

  • @justvinceb4427
    @justvinceb4427 7 лет назад +1

    TSP states you have to arrive back at original position. This vid does not take that into consideration. Plus, how can the distance from A, D, C be 33 and the distance from A, B, C be 34 and the distance from A, E, C be greater than them both at 45 when the triangles have 90 degree angles?

    • @alkestos
      @alkestos 2 года назад

      If it’s on a curved surface.

  • @parvanehshafiei1
    @parvanehshafiei1 10 лет назад

    Nice video..tnx..there is some problem at minute 3..for the problem the path is BE,EA,AD..u have written BE,ED and AD

  • @ThiNguyen-gd9ny
    @ThiNguyen-gd9ny 8 лет назад

    hello...
    how to write programs using HEAPSORT to
    solve Travelling Salesman Problem - TSP on programming language C ...
    and I need code :(
    im form Viet Nam...thanks

  • @benpct5555
    @benpct5555 5 лет назад +1

    You can hear him laughing when talking about the karma at 4.46.

  • @rainaw0924
    @rainaw0924 10 лет назад +2

    Why the lower bound is 62? Why add 18 and 19?

    • @yukaizhong4889
      @yukaizhong4889 10 лет назад

      You will know!

    • @rainaw0924
      @rainaw0924 10 лет назад

      Yukai Zhong Haha, I am solving the NP-Complete problem!

    • @yukaizhong4889
      @yukaizhong4889 10 лет назад

      You are the boss!

    • @rainaw0924
      @rainaw0924 10 лет назад

      Yukai Zhong you bet!

    • @edwardliu7203
      @edwardliu7203 10 лет назад +1

      Because the graph must have at least two paths, there must be a circle in the graph. And you have to to reach C via the shortest path, so we have to add 18 and 19.

  • @lerneninverschiedenenforme7513
    @lerneninverschiedenenforme7513 7 лет назад

    I think you should've made clear why you could travel a node twice, when thinking of an upper bound

  • @vocdex
    @vocdex 8 лет назад

    Could someone tell me whether or not ABC and ADC are triangles??? 16+18

    • @krzysztofbracha6852
      @krzysztofbracha6852 8 лет назад

      None of them are triangles, it is only a simple representation and the distances in the draft are not accurate.

  • @mayonesacosmica
    @mayonesacosmica 10 лет назад +5

    "there isn't really an algorithm for giving you the best route"
    How about brute force, held-Karp, branch-and-bound?

    • @PflanzenChirurg
      @PflanzenChirurg 5 лет назад +3

      Bruteforce is not the right way without harming alot due to the needed (energy/time). Think in a big scale

    • @jamilhneini1002
      @jamilhneini1002 5 лет назад

      @@PflanzenChirurg It might be feasible if the interval is small enough

    • @PflanzenChirurg
      @PflanzenChirurg 5 лет назад

      @@jamilhneini1002 logic is the only Thing you can apply to improve. Not any CPU Performance

    • @jamilhneini1002
      @jamilhneini1002 5 лет назад

      @@PflanzenChirurg reducing the possible values to make it feasible is a very logical way

    • @PflanzenChirurg
      @PflanzenChirurg 5 лет назад +2

      @@jamilhneini1002 nope, not really. You Need a complete logic to make it feasible ;)

  • @danamuise4117
    @danamuise4117 10 лет назад +15

    he solved the unsolvable! (not)

  • @sancheztorres8724
    @sancheztorres8724 9 лет назад +3

    Thank you for your great explanation!!!

  • @jcwondrous
    @jcwondrous 11 лет назад

    Is this a typical tsp? cos first it is not complete and second there is no triangular property (such as ABE)

  • @memyselfiyo_its_me_lmaooo8769
    @memyselfiyo_its_me_lmaooo8769 5 лет назад

    Man ur amazing .My teacher is trash , wish i had someone like u

  • @Guderian444
    @Guderian444 Год назад

    Right answer is 76, I dont understand, what is 62 in this case

  • @amaliehansen8199
    @amaliehansen8199 8 лет назад +1

    Does anyone know, who got up with the idea of "the optimal TSP-route"?? :-)

  • @sharzin1
    @sharzin1 10 лет назад +8

    TSP can't visit repeated vertex (not even to go back to the "start vertex") You are talking BS!

    • @Root_boy
      @Root_boy 10 лет назад

      Irrelevant. The shortest path from A to C has to be

    • @MartinFracker
      @MartinFracker 10 лет назад +7

      It's very well known that in the Travelling salesman problem, the salesman starts in a "home" city, visits all the cities on his agenda exactly once and returns to his home city. I'm afraid you are the one speaking BS.

    • @Root_boy
      @Root_boy 10 лет назад

      There's no need to be rude, especially when you're wrong it can make you seem a bit ignorant.
      It's simple deductive reasoning which holds true within, but not limited to, the rules of the traveling salesman problem.
      Given any route, the shortest route is shorter or equal to that route. That's all I'm stating.
      It is also well known that, within the traveling salesman problem, one may assume that any node can be connected to any node. Therefore, I suggest taking the shortest route from A to C. This is of course the direct route which the author of the video neglected to draw but can be assumed to be less than or equal to the indirect route through B.

    • @MartinFracker
      @MartinFracker 10 лет назад +2

      Rajiv Krijnen I'm not sure if you thought I was replying to you, but just in case: I was replying to Farzin Shargh

    • @Root_boy
      @Root_boy 10 лет назад +2

      Well that kind of makes sense. I was struggling to see the relevance of your comment in relation to mine. My bad.

  • @justsomeguy-u6p
    @justsomeguy-u6p 12 лет назад

    Seeing how to do this on a table would be awesome? :)
    Thanks for your help!

  • @amitchauhan2954
    @amitchauhan2954 9 лет назад

    Can Someone Explain Why 18 and 19 came

  • @TheSHJGaming
    @TheSHJGaming 8 лет назад

    Why can't you use Dijkstra's shortest path algorithm?

  • @aquawoelfly
    @aquawoelfly 8 лет назад

    adcbe... avoid 23... since all the outside routes are shorter then inside routes.

  • @samfraser1994
    @samfraser1994 11 лет назад

    It's not certain that 62 works, whereas 76 definitely works

  • @wayn0uk
    @wayn0uk 5 лет назад

    i am sorry but why isn't the lowest also the optimal?

  • @shapekam7243
    @shapekam7243 9 лет назад

    you only show that how to solve the travelling salesman problem but why you doesn't apply methods like dynamic programming and greedy ??

    • @nour.bs9144
      @nour.bs9144 8 лет назад

      I'm doing an assignment of TSP... I have to find a (non-evolutionary algorithm ) to solve the problem... do u know wt types of algorithms that are not belong to the evolutionary algorithm class? and does the dynamic algorithms consider as a non- evolutionary algorithms?

  • @LordDijon
    @LordDijon 12 лет назад

    Can you please make a video on nearest neighbor algorithm on a table pleaseee? :) P.S your videos are awesome

  • @shum8927
    @shum8927 11 лет назад

    the lower bound isn't 60???

  • @TheGreatBunghole16
    @TheGreatBunghole16 7 лет назад

    Mic is muffled, can't understand without distraction, skipping. :(

  • @james04gr
    @james04gr 10 лет назад +1

    dude u r wrong...!!!! Y u connected C with both 18 and 19?????

  • @imadsaddik
    @imadsaddik Год назад

    Thank you

  • @umeshmalhotra4617
    @umeshmalhotra4617 8 лет назад

    What is the exact answer?

    • @doll49
      @doll49 4 года назад +1

      What kind of question is this?(lol) The initial question asks for the lower and upper bounds and that's literally how he ends the problem.. by telling you those answers. Did you maybe confuse what the question is asking?

  • @AnishKaranPhotography
    @AnishKaranPhotography 12 лет назад

    helped me a lot...thanks so much

  • @ScottiStudios
    @ScottiStudios 8 лет назад +2

    Erm, how it C>D shorter than C>E? I know you have 19 for C>D versus 23 for C>E, this is clearly wrong. C>D, (the hypotenuse, as this looks like a right angled tri), should ofcourse be longer than the other two sides. I think you may have made a calculation error...

    • @MrHard2beet
      @MrHard2beet 8 лет назад +2

      the right angled triangles are just for visualisation, if the lengths of the sides were drawn correctly the triangles would not be right angled.

  • @jadejapruthvirajsinh3567
    @jadejapruthvirajsinh3567 5 лет назад +1

    saav khotu chhe. aam kai thay j ny. khotino

  • @mohben2216
    @mohben2216 4 года назад

    le graphe est en contradiction avec la géométrie euclidienne

    • @alkestos
      @alkestos 2 года назад

      Yeah, the surface must be curved but it’s somewhat irrelevant to the actual point. It’s just there for a visual help, not to present an actual scale image of the problem.

  • @mattdathew2794
    @mattdathew2794 10 лет назад

    thanks

  • @Kylirr
    @Kylirr 7 лет назад +1

    >konnichiwag1

  • @TurtleRox95
    @TurtleRox95 12 лет назад

    AWESOME!

  • @yogadharshiniu3054
    @yogadharshiniu3054 6 лет назад

    good 😊

  • @teodorkolev9510
    @teodorkolev9510 11 лет назад

    thank you!!!

  • @ajit5593
    @ajit5593 11 лет назад

    Thanks for that :D

  • @RaD1K4l
    @RaD1K4l 12 лет назад

    Anyone else get 75?..

  • @frenchpicker
    @frenchpicker 6 лет назад

    Right from the start ....closest neighbor of A is E not D. Your numbers are out of proportion

  • @KiranBV
    @KiranBV 12 лет назад

    konichiwagwan!

  • @ahmedboghdady1407
    @ahmedboghdady1407 11 лет назад

    thnxxxxxxxxxxxxxxxx

  • @superhero2316
    @superhero2316 7 лет назад

    koiln chatpa

  • @fakjbf
    @fakjbf 10 лет назад

    Anyone notice that in some of the triangles the legs are longer the hypotenuse? You broke geometry.........

    • @ssaz5
      @ssaz5 10 лет назад +5

      That was a graph bro. A graph from graph theory. It was not a geometrical shape. It was several nodes connected via pointers.

  • @punitkoushal5428
    @punitkoushal5428 6 лет назад

    doobmar

  • @gautambarua6056
    @gautambarua6056 9 лет назад

    waste vdo

  • @wayn0uk
    @wayn0uk 5 лет назад

    Stop with mouth sounds.

  • @guddubhagat1759
    @guddubhagat1759 8 лет назад

    the fuckkk !!!!! 18+19 is 62????????????? really??????????? at 3:24 minutes.

    • @__jay__
      @__jay__ 8 лет назад +8

      +guddu bhagat wouldn't it be nice if you try a little hard to keep up with things he added 18 and 19 to 25 which gives 62 😃

    • @ayushkotiyal4357
      @ayushkotiyal4357 7 лет назад +1

      guddu chutiya sala

  • @shawnlapuz9472
    @shawnlapuz9472 8 лет назад +1

    your hypotenuse cannot be shorter(lower value) than either side of the triangle; therefore your solution is invalid.

    • @JeffThompsonAtlanta
      @JeffThompsonAtlanta 8 лет назад +13

      +Shawn Lapuz Those are not triangles, it's a simplified graph.