7. In binary classification context, let p represent probability of class label 1, which imply that 1 -p represents probability of class label 0. Logistic function is the cumulative distribution function of logistic distribution, which maps a real number z to the open interval (0, 1): e2 It is easy to see that when z -00, function p(z) 0, and as z 00, function p(z) 1. Show that the inverse of logistic function is the logit function: z(p) = In p 1-p Logit link function is commonly used as a link function in binary classification. To see the function of the link function, suppose we use linear model to represent the unobserved quantity z. That is z = A)+A , for example. In such setup, z represents a log odds ratio that is being modeled with a linear model. Therefore, z -oo implies p 0 (predict class label 0) and, similarly, z oo implies p 1 (predict class label 1)