4. Logistic Regression

Linear regression is useful for data with linear relations or applications for which a first-order approximation is adequate. There are many applications for which linear regression is not appropriate or optimal. Because the range of the linear model in Equation 3.3 is all of $\mathbb{R}$, using linear regression for data with continuous outcomes in $(0,1)$ or binary outcomes in {0,1} may not be appropriate. Logistic regression (LR) is an alternative regression technique naturally suited to such data.