BayesianNetworks

Поделиться
HTML-код
  • Опубликовано: 9 ноя 2024

Комментарии • 27

  • @fiacobelli
    @fiacobelli  6 лет назад +26

    YES, many of you noticed... The exercise assumes that both John and Mary called. Sorry about that. if Mary hadn't called, I should have used m=f throughout.

    • @Ray-eb5cy
      @Ray-eb5cy 3 года назад

      I was scared for a second cause I thought I really understood up until then. I have a question though, let's say we weren't concerned about John nor Mary and we simply wanted to know the probability of P(B| E, A) do we still need to compute the probabilities involving John and Mary or would we apply variable elimination to remove them?

  • @mireazma
    @mireazma 7 лет назад +11

    Thanks for the clear explanations. I finally understand how to calculate probabilities in bayesian networks. I hope you don't mind but please use a consistent capitalization.

  • @moonman239
    @moonman239 Год назад +1

    If it helps, consider a simpler example:
    Suppose John and Mary have instead agreed to watch your house and then call you if they see that someone has broken in. If someone breaks in, John has a 70% chance of calling you, but he also has a 40% chance of calling you if a door-to-door salesman shows up to your door. If someone breaks in, Mary will have a 60% chance of calling you, but a 30% chance of reporting a door-to-door salesman. The probability of someone breaking into your house is 0.1% and the probability of a door-to-door salesman ringing your door is 10%.
    Both people will only call if they think you've been burgled, but we don't know when, only that they will call within a given amount of time after the person is spotted.
    First John calls you, so the probability of the call being due to a burglar is the probability of there being a burglar (0.1%), times the probability of John calling given there's a burglar (70%), divided by the probability of John calling you (which is the probability of John calling given there's a burglar + the probability of John calling given there's a salesman), or 0.1% * 70% / (0.1% * 70% + 10% * 40%) = 1.7%.
    Then Mary calls, and the probability you've actually been burgled is 1.7% * 60% / (1.7% * 60% + 10% * 30%) = 25%

  • @rasikapurohit
    @rasikapurohit 5 лет назад +6

    truly I wasted my tuition fee where a robot explained me to this. Thanks a ton!

    • @user-rf4vc7mt4d
      @user-rf4vc7mt4d 3 года назад

      im studying this before our prof started this topic. i have a project on it and i just know im not gonna learn anything from him

  • @iamtechboy3298
    @iamtechboy3298 6 лет назад +1

    WOW! Super amazing way. For this first time i'm knowing what that hack really Bayesian Networks is. Thanks, Long Live. Hats Off.

  • @meerageorge9656
    @meerageorge9656 8 лет назад +7

    Here at 14:02, Mary doesn't call is given in question. My understanding is that we have to evaluate for m=f then but it is taken as m=t. Kindly clarify the case.

    • @mireazma
      @mireazma 7 лет назад +2

      He actually talks about John calling and Mary not calling, on two occasions, both in which it's written m=t.
      I'm guessing the probability that we should evaluate with m=f instead of m = t, is 0.9 :)

  • @Igor-rd2fb
    @Igor-rd2fb Год назад +1

    4:40 How did they calculate P(A|B,E)?

  • @RoastedTuberLp
    @RoastedTuberLp 5 лет назад +3

    Holy shit thanks dude, you explain it in such an understandable way!

  • @AlanRickman2
    @AlanRickman2 5 лет назад +1

    Excellent explanation.

  • @mahnooraftab2220
    @mahnooraftab2220 2 года назад

    Hi, regarding to the conditional independence, we say that J and M are independent given A. But my question is that J and M are independent given B?
    Help will be really appreciated.
    Great video.

  • @tuongnguyen9391
    @tuongnguyen9391 4 года назад

    wow thank you so much, but what playlist of yours does this video belong to ?
    I am really feel grateful

  • @taggebagge
    @taggebagge 5 лет назад +1

    The usage of the names Mary and John will forever be ruined in my mind due to the video with that Indian fellow who explains the various usages of the word "fuck". Great vid though, thankss.

    • @AyushMo
      @AyushMo 3 года назад

      haha, curious which video you're referring to(although yes this was a year ago so I know I'm pushing my luck in expecting a reply but meh)

    • @taggebagge
      @taggebagge 3 года назад

      @@AyushMo Nah, you are certainly not pushing your luck in this regard, this one it is, all the best man:
      ruclips.net/video/Bdgi6PAtH1Q/видео.html

    • @AyushMo
      @AyushMo 3 года назад

      @@taggebagge Haha damn, thank you so much

    • @taggebagge
      @taggebagge 3 года назад +1

      @@AyushMo You are most welcome.

  • @qamariqbal2018
    @qamariqbal2018 Год назад

    I think I should have a PhD in math and then find out how should I respond to this alarm. I will let burglar to take away everything and leave me alone

  • @Claxiux
    @Claxiux 6 лет назад

    thank you for your help, these videos are very valuable!

  • @ifyonye6842
    @ifyonye6842 4 года назад

    Can i get your email to discuss a challenge i have with Bayesian Network?

  • @bakarandguladze
    @bakarandguladze 6 лет назад

    Thank you, mate! very helpful

  • @usmanabbas7
    @usmanabbas7 5 лет назад

    Thank you so much Sir :)

  • @navjotkaurbhamrah8233
    @navjotkaurbhamrah8233 6 лет назад

    This video really helped me. Thank you !!

  • @joeleepee
    @joeleepee 11 лет назад

    Great explanation!would u teach us about learning of bayesian network?TKS!

  • @thanhnguyenquang1513
    @thanhnguyenquang1513 3 года назад +1

    the sound is not good I am quite disappointed