Chapter 6 The Linear Model II: Logistic Regression
In the previous chapter, we introduced the linear model, and showed how we can use it to model continuous dependent variables (\(Y\)), using a combination of both continuous and categorical indepdent variables (\(X\)).
In this chapter we will discuss expanding our toolkit to use a different type of regression, logistic regression, to model categorical dependent variables (\(Y\)). For now, we only consider binary dependent variables (such as Yes
/No
Decisions, True
/False
classifications, etc), although there are also extensions to categorical dependent variable with multiple levels (e.g., multinomial regression).
The learning objectives for this chapter are:
- Readers should understand and be able to use logistic regression to estimate categorical dependent variables (e.g., to perform classification).