(ML 14.11) Viterbi algorithm (part 1)

แชร์
ฝัง
  • เผยแพร่เมื่อ 27 พ.ย. 2024

ความคิดเห็น • 26

  • @caseyli5580
    @caseyli5580 5 ปีที่แล้ว

    way better than the terrible jumbled mess that got given to me in class slides. super intuitive and clear. thank you!

  • @wavegeekman
    @wavegeekman 7 ปีที่แล้ว

    Thank you!. These videos helped me to understand. As you pointed out, most of the textbooks have very enigmatic explanations of all this.

  • @RoozbehSanaei
    @RoozbehSanaei 8 ปีที่แล้ว +8

    I like his videos, probably the most intuitive way of teaching HMMs I found so far, but there are just too many distractions and interruptions that can be confusing.

  • @ninosawbrzostowiecki1892
    @ninosawbrzostowiecki1892 9 ปีที่แล้ว +21

    Hey Sal, your voice changed, but your manner of speech remained the same. Also, I noticed your channel name changed... what's the deal? Did you get a haircut or something?

    • @ernie2111
      @ernie2111 9 ปีที่แล้ว +1

      Ninosław Ciszewski lol

    • @haoguoxuan411
      @haoguoxuan411 8 ปีที่แล้ว +4

      this is not Sal.

    • @BinuJasim
      @BinuJasim 3 ปีที่แล้ว

      @@haoguoxuan411 I am sure that was sarcasm. Probably Sal was inspired by MathematicalMonk to do his videos in this style.

  • @shiningyrlife
    @shiningyrlife 12 ปีที่แล้ว

    love your videos. You saving me from failing my bio infomatic course

  • @mulew5353
    @mulew5353 11 ปีที่แล้ว +15

    can you upload video that describe the baum welch algorithm please?

  • @swallace21
    @swallace21 7 ปีที่แล้ว +3

    Good explanation Math wise. But seriously if you want to help bridge that gap for some people please give a real example. Talking only in math is like asking a native english speaker to understand a new concept when you only talk about it in spanish.

  • @alecvan7143
    @alecvan7143 5 ปีที่แล้ว

    Great videos!!

  • @nasimsonboli3712
    @nasimsonboli3712 7 ปีที่แล้ว +1

    I have a question, if we continue multiplying the initial probability by the transition probability, because of the property of statistical equilibrium, we finally get to the stationary state, right? Can we conclude the same thing for these kinds of Markov Chains as well?

  • @adhoc3018
    @adhoc3018 3 ปีที่แล้ว

    4:29 I see it now, thanks!

  • @chogo888
    @chogo888 13 ปีที่แล้ว +1

    i love the viterbi algorithm!

    • @TheSubwayBeast
      @TheSubwayBeast 3 ปีที่แล้ว

      Please explain it to me then guy from 10 years ago.

  • @TommyCarstensen
    @TommyCarstensen 11 ปีที่แล้ว +1

    Is there a video with an example of the application of the algorithm to a real life problem?

  • @WilliamWangYang
    @WilliamWangYang 12 ปีที่แล้ว +1

    One equation, many lines of code.

  • @riccardoandreetta9520
    @riccardoandreetta9520 8 ปีที่แล้ว +10

    it is difficult to understand it without an example as you go on explaining the maths about it

  • @danishamjad5807
    @danishamjad5807 8 ปีที่แล้ว +3

    At 7:30, Why you used 1:k-1 instead of 1:k?

    • @baravi2005
      @baravi2005 8 ปีที่แล้ว +3

      Basically we are trying to find what's the optimal path (i.e. the set of k-1 states) that lead to z_k= s_j (we should assume that we are given the state value=s_j at z_k)
      You should also watch the second part

    • @danishamjad5807
      @danishamjad5807 8 ปีที่แล้ว +1

      I got it after watching second part. Thanks BTW. :)

  • @MotleyVideos
    @MotleyVideos 11 ปีที่แล้ว

    mathematicalmonk,I'm following most of your video.Thanks for the great work.Could you tell me the name of the draw to text tool you used that you think might have been built on the basis of hmm.