Logistic regression in R

Поделиться
HTML-код
  • Опубликовано: 2 фев 2025

Комментарии • 26

  • @EquitableEquations
    @EquitableEquations  Год назад +4

    You can find materials supporting this vid (and others) at github.com/equitable-equations/youtube.

    • @muqadasjaved7751
      @muqadasjaved7751 10 месяцев назад

      Can you please do multivariate detrended fluctuations analysis or multifractal detrended fluctuations analysis in rstudio?

  • @МаріяМуравйова
    @МаріяМуравйова Год назад +3

    omg thank you so much, I did almost the whole data analysis course at uni based on your videos😁 amazingly well explained and usable🙏

  • @romanvasiura6705
    @romanvasiura6705 Год назад +1

    Amazing, thank you for this video lesson!

  • @ahmedelkholy1323
    @ahmedelkholy1323 Год назад +1

    thanks a lot always adding something to me

  • @weberb2781
    @weberb2781 Год назад

    Hope you talk about the {fixest} package, it is very convenient to use this package for regression.

  • @badermuteb4552
    @badermuteb4552 Год назад +1

    how did you simulate the data and load it as csv? sorry for the newbie question

    • @EquitableEquations
      @EquitableEquations  Год назад

      Hi! I'm planning on making vids about simulating data later in the year. You might start by googling "inverse CDF method". You can write a data set to a file with write_csv() and read it back with read_csv(), both from the {readr} package.

  • @leoniepfeifer4091
    @leoniepfeifer4091 Год назад

    Hi Thanks for the great video!!Do i need to mark the binary variable as binary? Because in R its recognized as "num"

  • @Ammarsays
    @Ammarsays Год назад +1

    Is it necessary to convert binary response variable into "0" and "1". Mine is "Yes" and "No". Is there a better way to convert it?

    • @EquitableEquations
      @EquitableEquations  Год назад +1

      The response doesn't have to be 0 and 1, but it does have to be a factor with 2 levels. Probably as.factor() is your friend here.

  • @djprometheus923
    @djprometheus923 Год назад

    the GOAT

  • @libbybendror
    @libbybendror Год назад +1

    your videos are the best! Any chance you will do MLM video please?

  • @AlbertoFCabreraCasillas
    @AlbertoFCabreraCasillas Год назад

    Nice explanation of logistic regression. Even your intentional mistakes created learning opportunities. In a future follow up , may I suggest sharing with the viewer tips as to how to make meaning of the parameters (e.g., reporting log of odds)

  • @flaviaoliveira5949
    @flaviaoliveira5949 8 месяцев назад +1

    what if there's more than 2 groups they can fall into?

    • @EquitableEquations
      @EquitableEquations  8 месяцев назад

      It's still possible to use logistic regression in that case (for instance with indicator variables), but in practice other techniques tend to be used instead.

  • @romanvasiura6705
    @romanvasiura6705 Год назад

    Thank you!

  • @RomitGhosh-n6g
    @RomitGhosh-n6g 11 месяцев назад

    Why do we write se = False

  • @hannukoistinen5329
    @hannukoistinen5329 10 месяцев назад

    But you don't do any testing, which is mandatory to give reliable results for your client!!!

  • @FatGuyInALittleCoat1
    @FatGuyInALittleCoat1 4 месяца назад

    Good day!
    What if your GPA data was overwhelmingly skewed towards one of the binomial outcomes?
    Example: out of 100 total plot points, you had 98 plot points to the upper right hand part of your sigmoid line, all above a 3.0 GPA, and were admitted students.
    However, you had only one on the lower left sigmoid line that wasn’t admitted
    And more confusing….one that was admitted on the upper left of the graph with a low GPA. Those X variables would point towards a high probability of not getting admitted and that’s what the model would predict. How is this interpreted?

  • @damirb6294
    @damirb6294 7 месяцев назад

    interpretation of the results missing