Multiple Linear Regression in Matrix Form

Поделиться
HTML-код
  • Опубликовано: 22 окт 2024

Комментарии • 34

  • @fishfeelpain7764
    @fishfeelpain7764 3 года назад +9

    You explained in 20min what my professor tried but failed to get across for 2 hours. Thank you!

  • @tors-a4808
    @tors-a4808 4 года назад +11

    Best explanation I've found online. Thanks!

  • @sharonchetia54
    @sharonchetia54 3 года назад +2

    So glad I found this. Thanks a lot Steve

  • @johnalbus1549
    @johnalbus1549 6 месяцев назад +1

    Amazing lecture

  • @futureliang1704
    @futureliang1704 2 года назад +1

    Explained so clear!!!

  • @jackgaydar8806
    @jackgaydar8806 3 года назад +1

    Excellent presentation - thank you very much.

  • @radharaghuraman4178
    @radharaghuraman4178 3 года назад +1

    This was fantastic!

  • @alg_poste-tech
    @alg_poste-tech 6 месяцев назад +1

    thank you

  • @hbalachandraprabhu7195
    @hbalachandraprabhu7195 2 года назад

    Wonderfully explained. Great job

  • @manzooranoori2536
    @manzooranoori2536 2 года назад

    simply WOW. Thank you so much.

  • @kaleabenyew819
    @kaleabenyew819 2 года назад

    absolutely helpful

  • @cardboardhearts3228
    @cardboardhearts3228 3 года назад

    thank you!!!!

  • @johncumbo1218
    @johncumbo1218 3 года назад

    Thank you for this! I hate econometrics a little less now!

  • @michellecheng6549
    @michellecheng6549 3 года назад

    Super helpful! Thanks a lot~^^

  • @richapandey7679
    @richapandey7679 3 года назад

    Hi Steve, Great work! And thanks for these videos. Quick question from me: the Beta vector that you defined when we had only one observation (10:58) versus for all the N obs (18:11) is different right? If yes, can I say, the one at 10:58 is a transpose of one at 18:11? Thanks.

    • @stevel2037
      @stevel2037  3 года назад

      Sure could. I don't make the distinction here since there are two vectors being multiplied.

    • @richapandey7679
      @richapandey7679 3 года назад

      ​@@stevel2037 Thanks Steve. Is the beta vector (at 10:58) a row vector or column? From your lecture, we wont be able to multiply x and beta vectors if they are both 1 by k+1 vectors. We can only multiple if beta is column vector or if beta vector is initialised as row vector then we transpose it.

    • @stevel2037
      @stevel2037  3 года назад +1

      @@richapandey7679 Nope - doesn't matter as long as the vectors are the same length.

    • @stevel2037
      @stevel2037  3 года назад +1

      You really need to be careful with matrices, not so much with vectors as the dot product of vectors can still commute (order doesn't matter).

    • @richapandey7679
      @richapandey7679 3 года назад

      @@stevel2037 Thanks.

  • @atlangkopane8906
    @atlangkopane8906 2 года назад

    Hi Steve, at 17:27 is the beta vector not supposed to contain beta hats instead of just beta?

    • @stevel2037
      @stevel2037  2 года назад

      Yes - PowerPoint wouldn’t let me put a hat and the vector symbol together, so I just used the vector symbol.

  • @Intellectual_House
    @Intellectual_House Год назад

    Do you have a video about how to calculate u the mean squared error
    I am searching how to get it's equation mathematically
    Hope you could help
    Thank you
    From Algeria 🇩🇿

    • @stevel2037
      @stevel2037  Год назад +1

      Here you go: ruclips.net/video/EefPbiF9YRs/видео.html

    • @Intellectual_House
      @Intellectual_House Год назад

      @@stevel2037 Thanks I'am little bit struggling with MSE when it's with multiple variable so if you could help me with that it will be super kind :)

  • @emircansac5373
    @emircansac5373 3 года назад

    which one is the next video guys?

  • @priyankabhagwat4609
    @priyankabhagwat4609 3 года назад

    What is K+1 term? to be specific what is exactly 'k' here?

    • @엠제이-d7c
      @엠제이-d7c 3 года назад +1

      I assume it is the number of independent variables in the model

    • @stevel2037
      @stevel2037  3 года назад +1

      @@엠제이-d7c it is actually one more than that - it is the number of beta parameters that need to be estimated (that’s why the “+1” is there - because of the intercept term B_0). K would be the number of independent variables.

    • @엠제이-d7c
      @엠제이-d7c 3 года назад

      @@stevel2037 wow, thanks!!!

  • @HL-iw1du
    @HL-iw1du 3 года назад

    Imagine being a vector.

  • @mogoggy8205
    @mogoggy8205 3 года назад

    Are all the betas the same or different

    • @stevel2037
      @stevel2037  3 года назад

      Generally they are each different since each x variable has a different partial effect on the y variable.