00:02
In a logistic regression model, not regressional, sorry, regression, there's a couple of key components and we're going to go through this problem and go through some true or false statements and explain them a bit.
00:36
So number one, it's a bernoulli distribution.
00:52
This is false.
00:55
The response variable in a logistic regression model follows a binomial distribution, not a bernoulli.
01:12
A bernoulli models a binary outcome, which would be like success or failure, but a logistic regression models their probability of success in a binomial distribution.
01:31
Number two, it can be any distribution over the interval which is zero to one.
01:35
This is also false.
01:51
The response variable in logistic regression follows a binomial distribution, which models probabilities between zero and one, but it is specifically designed for binary outcomes, zero or one.
02:11
Three, it can be any distribution from the exponential family.
02:30
This is false.
02:33
Logistic regression models the response variable using a binomial distribution, which is a specific member of the exponential family.
02:41
It's not applicable to any distribution from the entire exponential family...