You could lime this video too: Another great video about logistic regression in JMP ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-9yN_yjGAJZE.htmlsi=jUwEZUDobBudE8AE
I have one doubt , is logistics regression = sigmoid(linear regression) or are there any other differences. Other than that amazing video, never found this much clarity.
@@pushkal8800 Yes, logistic regression is essentially a combination of linear regression followed by the application of a sigmoid function. However, while this captures the essence of how logistic regression models the relationship between the independent variables and the dependent variable, there are a few more nuances that distinguish logistic regression from simply applying a sigmoid function to linear regression.
13:49 is a mistake I think. odds ratio of 1.04 does not mean that the probability increases by 1.04 times. An odds ratio of 1.04 means that for a one-unit increase in the independent variable, the odds of the event happening (in the context of a binary outcome) are 1.04 times higher. It does not directly relate to a specific change in the probability of the event occurring. To understand the impact on probabilities, you would need to convert the odds ratio back to probabilities using the logistic function. The mistake is in confusing odds and probabilities.
Too many thanks. It's very well explained and understood. May you help and also make a detailed video on ordinal logistic regression and multinomial logistic regression, explaining the different equations used under each step by step like you have done under binary logistic regression? Thanks
I feel like the use of the word dichotomous was simply to sound cool. Considering logistic regression is often used in computer programs, I think binary would have been much more suitable of a term. Feel free to educate me on why dichotomous was used if it is a better term.
I think dichotomous is used to align with the existing and established name of a type of categorical variable in statistics in general; in this case, a dichotomous variable/data. While logistic regression may be often used in computer programs, its application actually transcends to varied fields especially in business and marketing, social science, and public health.
0.28,o.03 are less than 0.05 but you said in our observation that no value is less than 0.05, so there is no independent variable with a significant influence. in the chi^2 test, while explaining the difference between model , you said dependent variables are used in the logistic function. while we use independent variable. Could you consider these suggestion?
If we use a binary logistic regression test, should the independent variable be made into two categories? Can't we analyze it if the independent variable has more than two categories? Especially if the two independent variables are in ordinal data form and the dependent variable is in nominal data form.
Wrong. In logistic regression, the response variable is an attribute variable that can be binary, nominal, or ordinary. you were only talking about the binary response variable.
She is not wrong, the explanation is as good as one can get in a short video, of course, the response variable is the dependent variable, and for the binary she is right, and that is what the video is all about.
Hi!! Try this one… you may like: Another great video about logistic regression in JMP ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-9yN_yjGAJZE.htmlsi=jUwEZUDobBudE8AE
![Logistic Regression [Simply explained]](/img/logo.svg){kind=logo}
