Material

🎥 Watch Logistic Regression

• Slides

Today's Goal

• Understand logistic regression
• Be able to fit logistic regression in R

Quiz

Q - Which of the following is the correct scenario to use a logistic regression?

a. when we have multiple categorical predictors

b. when we have a mix of numeric and categorical predictors

c. when we have a count response variable

d. when we have a binary response variable

Quiz

Q - Which of the following is the correct scenario to use a logistic regression?

a. when we have multiple categorical predictors

b. when we have a mix of numeric and categorical predictors

c. when we have a count response variable

d. when we have a binary response variable

Quiz

Q - Which of the following is not correct for a logistic regression? Select all that apply.

a. it is a non-linear model

b. it uses a logit link

c. logit is a function of a binary response $y$

d. in order to fit the logistic regression in R, we define family = "binomial"

Quiz

Q - Which of the following is not correct for a logistic regression? Select all that apply.

a. it is a non-linear model $\to$ linear model with $logit\left(p\right)$ as a response (Generalized linear model)

b. it uses a logit link

c. logit is a function of a binary response $y$ $\to$ it is a function of the success probability $p$ of $y$

d. in order to fit the logistic regression in R, we define family = "binomial"

Quiz

Q - Write out a logistic model using parameters.

$$\log \left(\frac{p}{1-p}\right)={\beta }_0+{\beta }_{1}x$$ No error term!

Let's Practice Together!

Go to AE 23: Logistic Regression

Bulletin

• Project peer review due Wednesday, June 15 at 11:59pm

• HW02, HW04 due Thursday, June 16 at 11:59pm

• Submit ae22 by Wednesday, June 15 at 9:29am

• Submit ae23 by Thursday, June 16 at 9:29am