Statistics 101: Logistic Regression, Estimating the Probability

Поделиться
HTML-код
  • Опубликовано: 1 дек 2024

Комментарии • 112

  • @marisolgarcia-cruz2589
    @marisolgarcia-cruz2589 2 года назад +4

    My tuition fees should go straight to you! You explain and deconstruct statistics with such efficiency, pedagogy and humanity. Honestly, I learned a lot more by listening to your videos than by attending 3 hours lectures that left me completely clueless once over. Really, a huge thank you!

  • @marcoventura9451
    @marcoventura9451 3 года назад

    I am amazed by the ability to explain in so a coincise and precise way.

    • @Janeilliams
      @Janeilliams 3 года назад

      can you explain , how the coeiffents are calculated ?

  • @texaspolygraph
    @texaspolygraph 7 лет назад +1

    Brandon is a great instructor. He makes this material so reachable.

  • @ariannawhite64
    @ariannawhite64 9 лет назад +21

    This series of videos was incredibly helpful while working on a data project for my graduate-level regression course. Even more helpful would be another series of videos summarizing how to interpret regression coefficients for regression equations using dummy-coding, weighted and unweighted effects coding, contrast coding, continuous predictors, and interactions with all of these. It gets so confusing!

  • @shantanu275
    @shantanu275 9 лет назад +1

    I watched all the videos for Logistic Regression. You are a phenomenal instructor. Hats off!!

  • @nonyabeeswax
    @nonyabeeswax 4 года назад +1

    This series is so underrated. Thank you Brandon! you have a gift, don't hide that in a bushel basket.

  • @Hazlo_Bob
    @Hazlo_Bob 2 года назад

    Teacher... I don't know who you are, but I truly love you and respect you. THANK YOU VERY MUCH!

  • @sureshm430
    @sureshm430 3 года назад

    THANKS is not enough.. i should say more than that, I started my day with logistic regression, planned to complete in 2 to 3 hours.. i had gone through lot of stuff which lead to confusion between logit and sigmoid function..... felt bad and sad, that i couldn't make it, then your stuff came, it is SAVIOUR.. You explained very clearly, detailed every thing in simpler manner. I ended my day with SMILE with your stuff.. APPRECIATE.. APPRECIATE.. APPRECIATE... Expecting stuff on Machine Learning Algorithms too.

  • @evansbatung8492
    @evansbatung8492 4 года назад

    I have been trying to solve for p (hat) for a week, and I found it in your video in under 3 minutes. Good job Brandon.

  • @sumesharora1
    @sumesharora1 5 лет назад +1

    Excellent videos on Linear Regression. Very nice slides and a clear explanation. Thank you, Brandon.

  • @abhijeetpandey5353
    @abhijeetpandey5353 2 года назад

    best content ever on logistic regression

  • @rihangweleserakalala2242
    @rihangweleserakalala2242 7 лет назад +5

    Your explanations are the best. Feeling more confident for my Stats test tomorrow. Thank you! :)

  • @tibi536
    @tibi536 6 лет назад +9

    Will you consider going through the MLE?
    Thank you for these videos, i can totally say that these videos are pure gold for me. As a machine learning student who is thinking about a phd in 1 year, these kind of videos gave me a great boost of my actual knowledge!

  • @4XLibelle
    @4XLibelle 7 лет назад

    Finally starting to get a handle on logistic regression--a calculus I have always appreciated so it's nice to begin to understand it (other than the algebra...). Looking forward to learning about the maximum likelihood. Much appreciation, Brandon.

  • @ariacube07
    @ariacube07 5 лет назад

    thank you for making these videos. i am binge watching your videos for my statistics exam. wish me luck

  • @_xcelpro
    @_xcelpro 3 года назад

    Wou!! Finally I got the concepts. Thank you so much.

    • @Janeilliams
      @Janeilliams 3 года назад

      can you explain , how the coeiffents are calculated ?

  • @SmartMrSteve
    @SmartMrSteve 5 лет назад

    If I give you more than one thumb-up, I want to give you as much as I can. Thanks for your video. Your statistics lecture is the best ever.

  • @august4633
    @august4633 6 лет назад +1

    Great video. Wish my professor would teach like this. It took us 4.5 hours to go through this topic in class AND it was more difficult to understand. Thank you so much.

  • @rajatbhosale8188
    @rajatbhosale8188 6 лет назад

    Brandon you are doing a great job. This is helping me to get hands on machine learning concepts. Thank you!

  • @thesource8990
    @thesource8990 5 лет назад

    A very comprehensible explanation of the logit model. Thank you very much.

  • @sunithamohanram5411
    @sunithamohanram5411 9 лет назад

    Very nice explanation.. I am watching all the videos of Brandon .. They are wonderful.Thank you so much Brandon ..

  • @rhrh100
    @rhrh100 8 лет назад

    I've watched a few of your videos. Top notch. Great job!

  • @CreekCat1
    @CreekCat1 9 лет назад +3

    I tried to take a stats class (online through my Univ.) and dropped the class. I just talked to another grad student in our Environmental Science program, and she too, dropped that class. We both are great students. Brandon, you are a spectacular instructor. I wish I had of known about your videos a year ago. Alls well. I have weird data to analyse or I won't get my degree. I'm glad one of my committee members is going to help me, but I'm going to keep watching your videos. Stats is a must know if you do any kind of science or read many research papers. You are great.

    • @BrandonFoltz
      @BrandonFoltz  9 лет назад

      ***** Thank you Yvette :) Very happy you find the channel helpful. It is why I do it. Hang in there you can do it. Perseverance wins the race.

  • @muneebahmed9280
    @muneebahmed9280 3 года назад

    Thanks for sharing, I appreciate your approach. keep it up!

  • @amank8549
    @amank8549 7 лет назад

    You have a very good and systematic approach!! Some of my doubts are cleared now!

  • @Lfmginc
    @Lfmginc 4 года назад

    Awesome video and overview. Thanks Brandon!

  • @anahitkhach
    @anahitkhach 8 лет назад +1

    Dear Brandon! Thank you very much for these videos! So clear & clever derivations. And I see how much work goes into making such quality videos. Great job! Can just relax and be sure you will get us out that jangle of math & stats :) Let know if we (your audience) can be of any help in making more of these videos? Thank you!

  • @ciprianp495
    @ciprianp495 7 лет назад

    this is wonderful! you make it so simple to understand. I honestly read books about logistic regression, all so complicated! Anyway, keep up the good work, you are amazing :)

  • @yashjain6372
    @yashjain6372 2 года назад

    Loved it.

  • @juandavidmunoz2781
    @juandavidmunoz2781 9 лет назад

    Your explanation is perfect.. Thank you so much for doing this.. It is great..

    • @Janeilliams
      @Janeilliams 3 года назад

      can you explain , how the coeiffents are calculated ?

  • @neilarmstrong5810
    @neilarmstrong5810 6 лет назад

    Thanks Brandon, really helpful! If this series were extended to Probit models, it would make this a complete series for introductory level Logistic Regression.

  • @manzurekhoda7013
    @manzurekhoda7013 4 года назад

    Thank you very much for such a good explanation

  • @appeiroon
    @appeiroon 6 лет назад +5

    What are the odds of you making a video about MLE? It would be cool to understand how to estimate those coefficients

  • @Max-my6rk
    @Max-my6rk 4 года назад +2

    gosh ur lecture made me love stats so much... who needs drugs if you have stats! I am in love!!!

  • @bobaisbomb3818
    @bobaisbomb3818 8 лет назад

    Well explained, very helpful. Thank you!

    • @Janeilliams
      @Janeilliams 3 года назад

      can you explain , how the coeiffents are calculated ?

  • @xudong5269
    @xudong5269 4 года назад

    brandon best G.O.A.T!

  • @zeinabhashemi9998
    @zeinabhashemi9998 6 лет назад

    you are an amazing teacher. tnx :)

  • @elianos1008
    @elianos1008 9 лет назад

    Perfect,keep it up Pro

  • @BrandonFoltz
    @BrandonFoltz  9 лет назад +14

    *NEW* Statistics 101: Logistic Regression, Estimating the Probability. Enjoy!

    • @chetanmootien6024
      @chetanmootien6024 8 лет назад +10

      +Brandon Foltz could you plz
      make a video about calculating maximum likelihood for the logit model ...??

    • @juankou9243
      @juankou9243 8 лет назад +2

      +Brandon Foltz Hi I have a question. Can you explain a bit more about how you get the coefficient and FICO score?

  • @mehraadi09
    @mehraadi09 4 года назад +1

    Would you like to go through the calculation behind MLE ? That'd would be extremely helpful 😍

  • @ralitsakostova8844
    @ralitsakostova8844 8 лет назад

    This is soooo helpful. Thank you!!!!

    • @Janeilliams
      @Janeilliams 3 года назад

      can you explain , how the coeiffents are calculated ?

  • @footballvideos4246
    @footballvideos4246 7 лет назад

    I skipped the LOGIT-lectures in my statistics course. This video helped me a great deal though!

  • @rawhit007
    @rawhit007 4 года назад

    This channel is damn good.

  • @georgepapadopoulos9916
    @georgepapadopoulos9916 4 года назад

    Love u man ❤️ good job

  • @MsMalleyV
    @MsMalleyV 9 лет назад +1

    Hi Brandon. I'm new to statistics and is trying to understand it. I am very glad to come across your videos! My question is which video would you recommend watching for a newbie like me that knows nothing about statistics?

    • @BrandonFoltz
      @BrandonFoltz  9 лет назад

      Everything begins here: ruclips.net/user/bcfoltzplaylists :) Are you in a stats class now?

    • @MsMalleyV
      @MsMalleyV 9 лет назад

      Yes. It's my first time in stats and I know nothing. I just started to learn about the levels of measurements. It has been very confusing.

    • @BrandonFoltz
      @BrandonFoltz  9 лет назад +1

      MsMalleyV In addition to my playlists I also recommend Khan Academy for very good explanations.

  • @TheJjonhk
    @TheJjonhk 4 года назад +2

    Hi! thanks so much for sharing this. Can I ask how you get the 3.289 score for the numerator in P hat? I plug in the same numbers but get 1.19
    Thank you so much!

  • @izzthewizz6
    @izzthewizz6 3 года назад

    You are a legend

  • @cristinacorreia4327
    @cristinacorreia4327 8 лет назад

    Brandon, Could you post something explaining Entropy? I have watched several of your videos and they are excellent. Thanks for sharing your knowledge!

  • @abhinavsiom
    @abhinavsiom 9 лет назад

    Dear Brandon, could you please advise the storage loaction of the datasets which you have referred in your videos..thanks,

  • @cjunsun
    @cjunsun 2 года назад

    May I know how to calculate the 95% confidence interval for the odds ratio? What type of distribution does odds ratio obey? Thank you.

  • @iheersinkable
    @iheersinkable 6 лет назад

    Thank you! You really helped a lot!

    • @Janeilliams
      @Janeilliams 3 года назад

      can you explain , how the coeiffents are calculated ?

  • @rabielangbam16
    @rabielangbam16 5 лет назад

    Thanks for the amazing videos really appreciate it

  • @erza4876
    @erza4876 4 года назад

    3:23, 4:36, how do you get the TABLEs... More interested in how to calculate the coef...

  • @Janeilliams
    @Janeilliams 3 года назад

    hey brandon when i tried out with same sets of data(same as yours , 15 obesrvations as it is) , the coieffents were different , cant figure it out and cant procedd it without figuring it out , any keys ?

  • @rrezonlajqii
    @rrezonlajqii 4 года назад

    Hi Brandon, thanks for the video (s)!
    If I understood correctly, if 1 is included in the C.I (be it upper or bottom level) that variable doesn’t have any impact on the odds, right? In this example 1 is not included so that's why we expect a change of 1.0147 change. Correct me if I'm wrong. Thanks!

  • @jdnr6204
    @jdnr6204 8 лет назад +3

    how did u calculate coefficients in the video?/ or how to calculate beta0 and beta1?

    • @kamleshkarki5761
      @kamleshkarki5761 8 лет назад

      Did you get the answer?

    • @jdnr6204
      @jdnr6204 8 лет назад

      KAMLESH KARKI yeah

    • @jdnr6204
      @jdnr6204 8 лет назад +1

      you need to use gradient descent method to calculate them

    • @dannuttle9005
      @dannuttle9005 7 лет назад +1

      Gradient descent is one algorithm; I believe there are others (which may arrive at a solution in a more computationally efficient way). Briefly, gradient descent is really just guessing. Try one set of coefficients, and then adjust one of them up or down to see which set better fits the data. Keep doing this until you find the best answer for that coefficient. Then try adjusting another coefficient. Keep doing this until you think you've converged onto the best answer, or something sufficiently close to the best answer. This makes logistic regression very different, in a sense, from linear regression, which has a single, deterministic, absolute best answer for a given data set.

  • @kwtan3814
    @kwtan3814 7 лет назад +2

    How do you get 3.289?

  • @RievenRast
    @RievenRast 8 лет назад

    I want to know how you calculated the intercept and the slope of the phat fucntion. What was on the x axis and y axis when calculating this?

  • @maricelk6868
    @maricelk6868 9 лет назад +2

    Hi. Thanxs for your great videos. I got a litle bit confused about the odd ratio in the output, an have a question about that. ( I think this is show as Exp(b) in SPSS.) Can this odd ratio be interpreted as : for each unit increase in your creditscore, the odds for getting aproved increase with 1,47 %?

    • @BrandonFoltz
      @BrandonFoltz  9 лет назад

      Maricel K Watch video 5 and let me know if you have any questions. I go over that in detail there. Thanks for watching!

    • @maricelk6868
      @maricelk6868 9 лет назад

      ***** Thanks! I got my answer in that video. Iæll keep watching your videos! Maricel

  • @chintanvyas1868
    @chintanvyas1868 8 лет назад +1

    can the p-hat be more than 100% and less than 0% ?
    Also what is "e" in equation?
    Thanks for putting up this video, very helpful in learning and If I know answer to above will help me apply this in my decision making process. thanks a tonne.

    • @Quitoss
      @Quitoss 4 года назад

      Yeah what do you substitute to the e?

    • @patricknylen3331
      @patricknylen3331 4 года назад

      I have the same question about the value for e

  • @mariocassara1285
    @mariocassara1285 8 лет назад

    Question: the FICO score of 720 gave us 3.289:1 odds of being accepted for the loan; but what is the reference score???? i mean, 3 times more likely to receive the loan compared to what score?
    Thanks!

  • @lowqchannel
    @lowqchannel 3 года назад

    what if the constant's p-value is more than 0.05 . how to interpret this

  • @STONE9523
    @STONE9523 5 лет назад

    Thanks for the video. Quick Q, how do you regress Y=b0+b1*x1? It's clear what series X1 is, but what is your Y series here?

  • @lowqchannel
    @lowqchannel 3 года назад

    can logistic regression be used for interval data ?

  • @amelianoor-oshiro4477
    @amelianoor-oshiro4477 7 лет назад

    Hi Brandon, long time viewer here! Can you explain whether the Odds Ratio of 1.0147 (and CI) as well as the FICO coefficient of 0.0146 are eerily alike? Is that a coincidence? Or can I just slap on a 1 to get an odds ratio from the coefficients? Thanks!

  • @lilac524
    @lilac524 9 лет назад

    Can odds ratio be negative? If not why would Stata produce negative dy/dx (marginal effects) for independent variables?

  • @sudipkhanal1438
    @sudipkhanal1438 5 лет назад

    Dear sir
    How can we calculate the beta 0 and B1 manually like simple linear equation.

  • @jaishivaram
    @jaishivaram 8 лет назад

    Hi Brandon,
    Suddenly had a doubt, why is comparison done by odds ratio (i.e using odds and not probability)? Will it be proper to do probability ratio?
    Is this mainly because of the logit function?

  • @lulubsc22
    @lulubsc22 4 года назад

    How about the equation if you use more than one predictor?

  • @Max-rs1kr
    @Max-rs1kr 3 года назад

    Thanks!

    • @Janeilliams
      @Janeilliams 3 года назад

      can you explain , how the coeiffents are calculated ?

  • @Skandawin78
    @Skandawin78 6 лет назад

    this is where You tube video which are promising suddenly head a road block. at 3:23 suddenly we are shown a bunch of results and not sure how to arrive at them. Atleast a high level input would help. I'm using R , please clarify.

    • @BrandonFoltz
      @BrandonFoltz  6 лет назад

      +gsrini did you start at the beginning of the playlist?

  • @paulegon552
    @paulegon552 5 лет назад

    Hi, I just want to ask how to know the goodness of fit of the model. Thanks

  • @roseb2105
    @roseb2105 7 лет назад

    can you make a video on how the coefficients are calculated?

    • @Janeilliams
      @Janeilliams 3 года назад

      did you got it , could you please explain it to me , or where you found ?

  • @basiliszag
    @basiliszag 3 года назад

    Thanks

    • @Janeilliams
      @Janeilliams 3 года назад

      can you explain , how the coeiffents are calculated ?

  • @nomearod
    @nomearod 7 лет назад +7

    This video still didn't answer how beta0 and beta1 are calculated.

    • @esdras1282
      @esdras1282 7 лет назад +14

      You're right, but in video 3 of this series he says that the coefficients are estimated using MLE (maximum likelihood estimation), which he is not going to describe in this series.

  • @satishvavilapalli24
    @satishvavilapalli24 6 лет назад

    why do we seperate probabulity and odds. im little bit confused der

  • @nikhilashodariya
    @nikhilashodariya 7 лет назад

    can you please share the data-set used in the video

  • @Janeilliams
    @Janeilliams 3 года назад

    can anyone explain , how the coeiffents are calculated ?

  • @AmirAli-ji1xm
    @AmirAli-ji1xm 6 лет назад

    But how we can calculate bita0 and bita 1?

  • @hopo2hopo
    @hopo2hopo 9 лет назад

    Please put the "click the next video" at least in the middle.

  • @RASTAcMOUSE
    @RASTAcMOUSE 5 лет назад

    My g

  • @roshan6317
    @roshan6317 7 лет назад

    written as natural log of odds ratio & in audio and in formula written as log of odds (not log of odds ratio) ............check again

  • @panagiotisgoulas8539
    @panagiotisgoulas8539 4 года назад +1

    You let so many things unexplained and even worse don't reply back to comments.
    NO explanation why if the confidence interval contains 1, then that means that increasing the credit score by 1 point does not have any effects on odds''?? I do realise 1 is an issue but rest values aren't.
    Similarly on previous video confusion between odds and odds ratio and why "the odds ratio for a variable in logistic regression represent how the odds change with a 1 unit
    increase in that variable holding all other variables constant''
    Elaborate please.

    • @Janeilliams
      @Janeilliams 3 года назад

      can you explain , how the coeiffents are calculated ?